Internal and external resources, tiredness and the subjective well-being of family caregivers of older adults: a case study from western Thailand, Southeast Asia

Abstract

This study examined the effects of several aspects of care-giving on the subjective well-being (SWB) of family caregivers of older adults. These aspects were comprised of external resources including social support, affection, and internal resources including confidence, positive appraisal and coping strategies, and tiredness. The level of happiness (0–10) was used as the gauge of SWB. The data were from a cross-sectional study of 270 primary family caregivers in western Thailand. Multiple regression models controlling for age, gender, kinship to care for recipients, self-rated health, duration and hours of care-giving were estimated. Both social support in terms of appreciation and admiration for care-giving, as opposed to such care being considered a caregivers’ duty, as well as inner strength derived from a positive attitude and self-perceived ability in care-giving were positively associated with SWB. Tiredness had a significant negative association with SWB. The implications of these findings for policy and intervention programs are discussed.

Introduction

Throughout the world, most informal caregivers are family members (Conner 2000; Knodel and Chayovan 2008; Novak 1997). Unfortunately, depression, poor perceived health, and an increased risk of mortality are well-documented among caregivers (Schulz and Beach 1999; Schulz et al. 1995). Their care-giving requires time and energy and is often given without financial compensation. It should, however, be noted that in some European countries such as the UK and Finland informal caregivers can receive cash benefits to replace lost income, linked to social protection coverage or as an acknowledgement of the work of care-giving. Additionally, care leave schemes that allow caring relatives to take some time off from gainful employment, to reduce their working hours or to provide care in the terminal phase of life, are available in a number of countries such as Germany and France (Spasova et al. 2018).

Care-giving is not always a negative experience or considered a burden, and there are positive aspects that can accrue to all family members (Mendez-Luck et al. 2008). Care-giving can be rewarding and can make a positive contribution to caregivers’ lives (Bertrand et al. 2012; Kuuppelomaki et al. 2004; Ribeiro and Paul 2008). The positive appraisal of the care-giving experience has been labelled as “positive aspects of care-giving” (PAC).

Research on family caregivers, however, usually focuses more on stress and feelings of being burdened (Carbonneau et al. 2010; Pinquart and Sorensen 2003) while research on PAC still lags behind the negative consequences of care-giving (Semiatin and O’Connor 2012). Most previous research has focused on the predictors of PAC (e.g. Lopez et al. 2005; Roff et al. 2004; Tang 2011) or negative outcomes such as the burden felt by caregivers (e.g. Contador et al. 2012; Lai and Thomson 2011). Much less research has been done on the associations of PAC and negative outcomes with SWB, and it has been suggested that more research on the correlates of SWB is needed (Pinquart and Sorensen 2003). There is also a call for research to examine whether a positive experience of gain in the care-giving role predicts better general well-being than a mere negative appraisal of care-giving (Kramer 1997).

In addition, previous research suggests that culture, including beliefs and values concerning care-giving, can influence the negative and positive appraisal of care-giving (Tang 2011). It has also been suggested that understanding the specific cultural values of target groups will be important for programme interventions. Most previous research has been done on Westerners, among whom, in general, individualism is a central cultural value. Very few studies have focused on other cultures, along the lines, for example, of Knight and Sayegh’s study of filial piety among East Asians (2010). There are also few studies addressing PAC in Southeast Asia where being connected to others is a central cultural value (e.g. Gray et al. 2016; Pankong et al. 2018; Yiengprugsawan et al. 2012).

Relevancy of Thailand: ageing population and its cultural context

Thailand is a middle income country in Southeast Asia that is currently experiencing rapid population ageing. The proportion of Thai people 60 years and older was 18% in 2018 and is projected to increase to 20% in 2021 and 32% by 2040 (National Economic and Social Development Board 2013). The proportion of older people is ranked second in the region after Singapore. Fertility in Thailand has declined very rapidly. Nowadays, a woman has an average of 1.6 children and that level is expected to decline further. Due to fertility decline and migration of adult–children to work elsewhere, care-giving by adult–children may be increasingly rare. This may especially be so in rapidly modernizing societies in the Asia region, where the traditional role of the family and especially children’s duty of care for parents (“filial piety”) may be weakening. Thus, there is a need for care-giving information to inform policies and interventions.

In Thailand, altruism and filial piety has traditionally been a norm, which is influenced by Buddhism, the religion of over 90% of Thais. Theravada Buddhism is said to influence people attitudes, thoughts and way of life (Knodel et al. 1987). Based on Buddhist teachings, care-giving to older parents is an obligation of children. Furthermore, the preference for one’s children as the main care providers when a parent is frail is widespread (Knodel et al. 2015). The norm of gratitude is not only for one’s parents but also for the others who have provided help or care for him or her. But, just as in many other cultures, Thailand is experiencing dramatic demographic, socioeconomic and cultural changes. There are an increasing number of spouses or other relatives who have become caregivers.

Due to fertility reduction and migration of children to work elsewhere, co-residence of people 60 years and older with children declined from 72.8% in 1994 to 54.7% in 2014, while those living with only their spouses increased from 11.6 to 19.0% during the same period. The long-term care system in Thailand has been established and based on the concept that the well-being of older persons can be attained through living with the family or community for the longest possible time. All of these factors render caregivers who include children, spouses, or relatives inadequate for the increasing care demand. Although the government is trying several potential solutions, for instance, community-based volunteers, these also have been insufficient to meet the increasing care demand (Foundation of Thai Gerontological Research and Development Institute and College of Population Studies 2013; Knodel et al. 2015). This situation may affect the well-being of both caregivers and care recipients.

Appraisal of strain or burden, appraisal of gain and well-being of caregivers

Previous research reveals that the appraisal of strain or burden and appraisal of gain (positive aspects of care-giving) in care-giving can be independent and have different impacts on well-being (Rapp and Chao 2000). Although “burden” is often studied as an outcome in its own right (Ostwald et al. 1999), it is also used as a predictor of other care-giving outcomes (Chappell and Reid 2002; Hughes et al. 1999). Burden is commonly understood to be composed of both objective and subjective components. Objective burden often refers to events and activities associated with negative care-giving experiences, while subjective burden is the emotional reaction of the caregiver such as worry, anxiety, and fatigue (Montgomery et al. 1985; Etters et al. 2008). Fatigue is an extreme tiredness resulting from mental or physical exertion or illness. Previous research shows that both subjective burden and objective burden were negatively related to SWB (Chappell and Reid 2002; Hughes et al. 1999). Thus, it is expected that caregivers who report higher burden will be more likely to have a lower level of SWB.

Concerning PAC, a meta-analysis done by Pinquart and Sorensen (2004) showed that PAC are strongly positively correlated with SWB. From previous studies, positive aspects of care-giving may be grouped into three domains: (1) the relationship between caregivers and care recipients; (2) social support or external resources; and (3) inner strength of caregivers. For the first domain, the data showed that better quality relationships with care recipients were associated with more satisfaction with care-giving (Lyonette and Yardley 2003). Caregivers who reported a high level of intimacy and love were less likely to show negative effects and reported more positive effects and overall psychological well-being (Braithwaite 1996). Thus, it is expected that those who received more affection from care recipients will be more likely to have higher SWB.

The relationship between caregivers and care recipients may create satisfaction in caring derived from the interpersonal dynamics between caregivers and care recipients (Nolan et al. 1996). Quality, interpersonal relationships can accrue benefits beyond those gained by caregivers and care recipients including broader social support or external resources. For example, care-giving can bring social recognition (Motenko 1989). Nolan et al. (1996) reported that recognition is important because it validates caregivers in terms of their competency. Caring can make caregivers feel appreciated (Noonan et al. 1996; Tarlow et al. 2004) and feel admired (Braithwaite 1996). Several studies have reported that the receipt of social support from others can reduce the negative effects of care-giving on physical health outcomes. Perception of support from the family and confidantes or friends has been linked to the caregivers’ physical and mental health (Li et al. 1997) and well-being (Chappell and Reid 2002). Therefore, it is expected that the caregivers who receive social support in terms of appreciation and admiration as well as other support from the community and the family will have greater SWB.

Another form of satisfaction of caring is derived primarily from the intrapersonal orientation of caregivers or internal resources. The feeling of accomplishment can lead to inner strength from things caregivers learn about themselves through care-giving. Feeling proud and competent boosts self-esteem and confidence in their ability to handle a crisis and can contribute to internal strength leading to fewer depressive symptoms (Kramer 1997; Pinquart and Sorensen 2003, 2004). Coping strategies, particularly religious coping methods can contribute to strength of caregivers. For instance, being rooted in Buddhist teachings, religious practices such as mindfulness and belief in good deeds (e.g. gratitude to parents by providing care) can lead to a positive appraisal (Gray et al. 2016; Pankong et al. 2018). Braithwaite (1996) found that a caregiver’s inner strengths were significantly related to enhanced psychological well-being. The more people can do to alleviate the hurt, stress, or suffering caused by a negative event or situation, the happier they will be. Thus, it is expected that stronger inner strength among caregivers will result in higher SWB.

To increase current knowledge about the PAC on family caregivers and to understand more about the complexity of care-giving, this study examines the effects of internal and external resources, as well as tiredness, on the SWB (i.e. happiness) among family caregivers who are providing care to their older dependent relatives. Ideally, results will support increased research and policy interest in the position of informal caregivers and the development of intervention programs to maintain this presumably very cost-effective care resource (Brouwer et al. 2005).

Methods

Data source

The data come from a survey of the Project of Demographic, Socioeconomic, Cultural and Long Term Care Surveillance for Well-being among the Thai Elderly in Kanchanaburi Province, Thailand, 2011. The survey was nested in a demographic surveillance site (DSS) in Kanchanaburi province, western Thailand. Kanchanaburi is Thailand’s third largest province in terms of area, bordering on Myanmar. The central town of Kanchanaburi is about 120 km from Bangkok. The DSS began in 2000, with a 92% individual response rate. It is an annual census of population and households in 100 sampling units in the province. All members age 15 years and older in the households were repeatedly interviewed in successive waves. A stratified systematic design was used for selecting villages in the rural areas and census blocks in the urban areas (Institute for Population and Social Research 2001). The selection of the 100 sampling units reflects the diversity in the province, comprising of 20 sampling unites in each of the following strata: (1) rice producing, (2) plantation crop producing, (3) uplands areas, (4) urban/semi urban (industrialized), and (5) mixed economy (Abas et al. 2013).

The 2011 Project is the first and only survey in Kanchanaburi DSS collecting data on caregivers to older adults. Eligible subjects were all those living in DSS households containing with at least 1 person 60 years and older on November 1, 2011. Caregivers were identified through screening questions on the disabilities of eligible subjects. Eligible respondents with a limitation on at least one of the basic activities of daily living (ADLs) (i.e. getting up from lying down, using toilet, bathing, dressing, washing face or brushing teeth, putting on shoes, grooming self and eating) or limitations on at least two of the instrumental activities of daily living (IADLs) (i.e. Taking bus or boat by own, taking medicine and counting change) were included in the study. These limitations are self-assessed. Only primary caregivers who lived in the same households as care recipients were interviewed. These interviews were conducted face-to-face. If there was more than one caregiver, “primary” caregivers were those who spent the most time on care-giving.

A consent form was provided to each subject. The interviewers read and explained the information on respondents’ right to abandon the interview or withdraw their participation at any time without any repercussion on the care of their older relatives or their own support services. Respondents were informed that their answers would be kept confidential and used only for research purposes. All respondents gave informed consent first verbally then signed the form. The study project was approved by the Committee for Research Ethics (Social Sciences), Mahidol University, Thailand (COA. No. 2012/063.0903).

Originally, there were 4345 older people and the 348 eligible caregivers (family, non-relatives and paid caregivers). Among these, 64 caregivers who reported unknown average hours of care-giving and 14 caregivers who were non-relatives and paid caregivers were deleted from the analysis. Thus, the final sample was 270 family caregivers. It should also be noted that there were no significant differences in demographics in the respondents who reported hours spent in care-giving and the respondents who were eliminated because they did not report hours spent in care-giving (Table 1).

Table 1.

Demographic and socioeconomic characteristics of participants in our sample (n = 270) and those with unknown hours of care-giving or non-sample (n = 64)

Characteristics of caregivers	%		Chi square	Asymp sig. (2 sided)
	Sample	Non-sample
Age			.471	.790
15–24	4.8	6.5
25–59	70.0	66.1
60 and over	25.2	27.4
Gender			1.295	.255
Male	25.2	32.3
Female	74.8	67.7
Kinship to care recipients			0.82	.960
Spouse	25.9	27.2
Children	53.3	54.8
Relatives (e.g. children in law and siblings)	20.8	21.0
Self-rated health			1.111	.292
Poor	20.4	14.5
Good	79.6	85.5

Measures

Dependent variable: subjective well-being

The level of happiness is used as the measure of SWB of caregivers. The happiness measure was based on a global question: “Taking all things together, the feeling of your happiness is at the level of…” Responses ranged from 0 (not happy at all) through 1, 2, … to 10 (happiest). Well-being research has commonly used a single-item instrument referring to happiness (e.g. Helliwell et al. 2018). One study showed that measuring happiness using a single item was reliable and valid since the answers had a high positive correlation with those provided by other happiness scales or inventories (Abdel-Khalek 2006). Additional advantages of using a single, global question can be found in George (2010).

Independent variables

Independent variables included factors related to negative aspects or burden and PAC. Tiredness has been used as a negative appraisal of informal caregiver burden (Broese van Groenou et al. 2013). It was found that higher tiredness was correlated with higher burden among informal caregivers (Peters et al. 2015).

For burden, a subjective measure of tiredness was used. It was based on the question “Do you feel tired of care-giving?” The response options were physically tired, mentally tired or physically and mentally tired.

PAC included external resources: perceived and actual social support from community and other family members, perceived affection from care recipients, and inner strength. Perceived social support for the caregiver was based on the question “How do people in the community view your care-giving?” The possible answers were: (1) They feel it’s your duty; (2) They feel sympathetic towards you; and (3) They feel appreciation and admiration for you.

Actual community support was based on the question “Have you ever received help in care-giving from the community?” This was coded as (1) yes and (0) no. The caregivers were also asked about actual support from any other family members: “Apart from you, do you have any relative helping you on care-giving?” The responses were coded (1) yes and (0) no. If yes, it was followed by the question how many helpers are there? and what per cent of the total work a helper (s) gave you.

Interpersonal relationship: perceived affection from care recipients

Perceived affection from care recipients was based on the question “During the last month, have older adults shown affection for you?” Responses were divided into three categories: rarely (never and a little), often (a lot) and always (very much).

Internal resource: inner strength

The question used in the survey was “During the last month, what were your views or feelings about the following statements?”

1. Confidence in my ability to provide care: “I have the ability to look after care recipients”.

2. Positive appraisal to provide care: “I feel good about looking after care recipients”.

The responses were combined into rarely (never and a little), often (a lot) and always (very much).

Buddhist coping strategies were based on two questions: “When you feel stress from care-giving, do you meditate or pray” and “When you feel stress from care-giving, do you accept the situation as a result of a poor deed from the past (karma)”. Response options were yes or no.

Control variables

They include caregivers’ personal characteristics which are age, gender, kinship to care recipients, health, length of care-giving and hours per day of care-giving (Chappell and Reid 2002; Hughes et al. 1999; Lopez et al. 2005; Tang 2011).

Analytical strategies

Multiple regression was used to explore factors affecting caregivers’ SWB. The dependent variable was the level of happiness, which was treated as a continuous variable. There were three statistical models. Model 1 represented the simple regression model in which happiness was regressed only on tiredness from care-giving. Model 2 included PAC. Model 3 included variables from Model 1 and 2 and control variables.

Results

Table 2 shows that the average age of caregivers was about 50 years. The majority of caregivers were females (75%), and more than half of the caregivers (53%) were children, followed by spouses (26%). Most of them (79%) reported good health. Among children, 78% were daughters. The average time spent caring per day was about six hours, and mean duration of care-giving was about 54 months.

Table 2.

Demographic and socioeconomic characteristics of caregivers and mean happiness

Characteristics of caregivers	%	Happiness
		Mean	SD
Age (Mean = 49.9, SD = 14.56, Min = 15, Max = 85)
15–24	4.8	8.08	1.98
25–59	70.0	7.66	2.08
60 and over	25.2	7.74	1.97
Gender
Male	25.2	7.75	1.84
Female	74.8	7.68	2.11
Kinship to care recipients
Spouse	25.9	7.66	2.11
Children	53.3	7.67	2.09
Relatives (e.g. children in law and siblings)	20.8	7.80	1.89
Self-rated health
Poor	20.4	6.96	2.09
Good	79.6	7.88	2.00
Hours per day of care-giving (Mean = 6.00, SD = 5.99, Min = 1, Max = 24)
0–4	57.4	7.76	2.10
5–10	25.2	7.44	2.04
More than 10	17.4	7.85	1.86
Length of care-giving (month) (Mean = 54.1, SD = 57.68, Min = 0, Max = 384)

n = 270

Considering the negative side of care-giving, about 41% of caregivers reported that they felt tired (Table 3). Most of them felt tired physically and mentally. Regarding the PAC, more than half of the caregivers (53%) reported that they often received affection from their charges. External resources were also important. Almost half of the caregivers reported that the community considered it to be a duty (46.7%). However, one-third of the caregivers (37%) felt appreciated and admired by their communities and they were the happiest group in this regard. About 20% of the caregivers received help from their communities. About 60% of the caregivers received help from other family members; among them, over 60% had 1 helper and 30% had 2 helpers. Regarding internal resources, about 10% reported that they rarely felt good about care-giving.

Table 3.

Negative and positive aspects of care-giving and mean happiness

Characteristics of caregivers	%	Happiness
		Mean	SD
Negative aspects
Being tired (perceived burden)
Yes	41.5	7.02	2.12
Physically tired	8.9	7.31	1.85
Mentally tired	9.3	7.12	2.49
Physically & mentally tired	23.3	6.87	2.08
No	58.5	8.17	1.85
Positive aspects
Interpersonal relationship
Perceived affection from care recipients
Rarely	18.9	7.18	2.26
Often	53.0	7.47	2.00
Always	28.1	8.47	1.76
External resource
Perceived social support from the community
It’s a duty	46.7	7.47	2.10
Sympathy	16.3	7.16	2.38
Appreciation and admiration	37.0	8.22	1.70
Received help from the community
Yes	20.7	7.98	1.99
No	79.3	7.62	2.06
Having at least one relative helper
Yes	59.6	7.56	2.24
No	40.4	7.79	1.90
Internal resource
Feeling good for looking after care recipients
Rarely	9.6	6.27	1.66
Often	56.3	7.46	1.95
Always	34.1	8.49	2.00
Perceived ability to look after care recipients
Rarely	40.4	7.32	2.14
Often	35.9	7.54	1.97
Always	23.7	8.58	1.73
Coping strategies
Meditation/praying
Yes	40.4	7.46	1.96
No	59.6	7.86	2.09
Acceptance of the situation
Yes	84.1	7.68	2.09
No	15.9	7.77	1.84

n = 270

Regarding religious coping strategies when they felt stress, most of the caregivers reported that they accepted the situation (84%). About 40% of caregivers practiced meditation or praying to relieve stress.

Bi-variate correlations for all variables included in the present study were calculated. All of the variables are weakly or moderately related to one another. (The results are not shown in the paper.)

According to Model 1, all types of feeling tired were statistically significant. The results reveal that those who felt tired physically and emotionally were the least happy (p < .001), followed by those who felt tired emotionally, and those who felt tired physically, respectively. This negative aspect of care-giving explained about 7% of the variation in happiness (Table 4).

Table 4.

Multiple regression analysis of happiness

	Model 1			Model 2			Model 3
	B	Std. Error	β	B	Std. Error	β	B	Std. Error	β
Negative aspects
Perceived feeling tired (Reference: No)
Tired physically	− .419	.216	− .117*				− .433	.214	− .121*
Tired emotionally	− .350	.142	− .149**				− .281	.137	− .120*
Tired physically and emotionally	− .324	.073	− .269***				− .254	.073	− .211**
Positive aspects
Interpersonal relationship
Perceived affection from care recipient
Always (Reference: Rarely, Often)				.425	.300	.094	.268	.299	.059
External resource
Perceived appreciation and admiration from the community (Reference: Duty, sympathy)				.583	.244	.138*	.700	.243	.166**
Received help from the community (Reference: No)				.017	.297	.003	.085	.301	.017
Received help from at least one relative helper (Reference: No)				.162	.244	.039	.141	.241	.034
Internal resource
Feeling good from looking after care recipient
Always (Reference: Rarely, Often)				.858	.247	.261**	.781	.278	.181**
Perceived ability to look after care recipient
Always (Reference: Rarely, Often)				1.004	.279	.209***	.871	.279	.181**
Religious coping (Meditation and/or praying)
Yes (Reference: No)				− .555	.238	− .133*	− .523	.237	− .126*
Acceptance of the situation
Yes (Reference: No)				.173	.322	.031	.378	.320	.068
Control variables
Age							.018	.011	.125
Gender (0 = Male, 1 = Female)							.185	.275	.039
Kinship to care recipient (Reference = child-in = law and siblings)
spouse							− .153	.413	− .033
Child							− .012	.294	− .003
Perceived good health (Reference = Poor)							.734	.298	.145*
Hours per day in care-giving							.009	.020	.025
Length of care-giving							.000	.002	.009
Adjusted R square			0.069		0.145			0.197

n = 270

*p < .05; **p < .01; ***p < .001

Model 2 showed that positive aspects explained about 14.5% of the variations in happiness. Perceived appreciation and admiration from the community, feeling good from looking after the care recipient and the perceived ability to look after the care recipient were positively and significantly associated with SWB. The other positive aspect variables were in the expected direction, but not statically significant. Unexpectedly, practicing meditation or praying had a negatively significant association with happiness.

Similar patterns to Models 1 and 2 were found in Model 3. There was, however, a weaker relationship between tiredness physically and emotionally, perceived ability to look after the care recipient and happiness (p < .01). Perceived appreciation and admiration from the community had a stronger effect on happiness (p < .01). None of the control variables had any significant effect except for health.

Table 5 shows reasons that caregivers reported for feeling mentally and physically tired. Almost 60% of the caregivers reported that they were physically and mentally tired due to the stubbornness of the care recipients. Many also reported that they did not have time for rest and for themselves. Lack of financial support for care-giving was also reported by about 36% of the caregivers.

Table 5.

Reasons for physically and mentally tired reported by caregivers

	Physically tired	Mentally tired	Mentally and Physically tired	Total
Care recipients
Being stubborn	16.7	76.0	57.1	52.7
Being offended easily	16.7	40.0	46.0	38.4
Self-centred and not reasonable	4.2	28.0	46.0	33.0
Get angry easily	16.7	36.0	39.7	33.9
Not cooperate with caregivers	8.3	28.0	38.1	29.5
Having conflict with care recipients	20.8	28.0	23.8	24.1
Caregivers
Not enough rest time	50.0	24.0	47.6	42.9
Do not have time for themselves	33.3	24.0	46.0	38.4
Lack of financial support for care-giving	16.7	36.0	42.9	35.7
Lack of encouragement when suffered from care-giving	8.3	28.0	27.0	23.2
Lack of knowledge of care recipients’ illness or health	12.5	24.0	25.4	22.3

n = 112

Discussion

As the increase in population ageing is a global phenomenon, the demand for informal care-giving is becoming an ever more important concern for researchers and policy makers. This study is one of few studies that explores the association of both PAC and negative outcomes with SWB (Pinquart and Sorensen 2003). Although this study is only based on Thailand, we are able to draw several conclusions considering the existing literature.

Our findings reveal that both negative and positive experiences in care-giving had significant associations with SWB of caregivers, when controlling for the potential confounding effects of sex, age, hours per day, length of care-giving, and health. Negative aspect (i.e. tiredness) had a negative association with SWB. Caregivers who felt tired often reported not enough time for rest and for themselves and lack of financial support. Thus, respite care by community volunteers to allow informal caregivers to take a break from caring tasks or cash allowances for family caregivers may help to reduce this negative aspect of care-giving.

For positive aspects, external and internal resources were positively related to SWB. Perceived social support (external resource) is important to family caregivers (Lai and Thomson 2011). Based on this study, emotional support in terms of appreciation and admiration is rather preferable than the traditional norm of the duty of care-giving. Interestingly, actual social support from the community is not associated with SWB. This finding may be because, based on the same data, most of the support is instrumental support, e.g. taking older people to hospitals or helping with housework. These tasks may be regarded as less important than emotional support. Additionally, having at least one relative helper does not increase SWB significantly. This may be due to the fact that, based on the same data, most of the caregivers perceived that they received help for less than half of the total care-giving task.

Another unexpected finding concerning a positive aspect is that caregivers who reported that care recipients show affection for them often had no significant difference in their SWB compared to those caregivers who did not receive affection from their care recipients. This is not consistent with previous research (Erlingsson et al. 2012; Kinney and Stephens 1989; Louderback 2000). In the case of children, filial piety is an ingrained value in the culture in Thailand and other countries of Asia. Respect for and caring for parents is a duty and obligation of adult children (Quah 2014; Sung 1998; Thanakwang et al. 2012). Perhaps, whether or not parents express affection to adult child caregivers, filial piety underpins adult child caregivers’ fulfilment of this cultural responsibility (Lai 2010). The finding for spouses may, however, be explained by the fact that affection from care recipients is often based on prior relationships that were established before the care recipient was in need of care. Due to a good relationship before the need for care, a caregiver may accept the challenging behaviours of a care recipient due to illness or a high level of dependency. This information on care of spouses was not collected in the survey used for this study.

With regard to internal resources, results here reveal that caregivers who had a higher level of inner strength due to feeling good about care-giving or to a perceived better ability in providing care had greater SWB. This finding is consistent with previous studies in Western countries (e.g. Braithwaite 1996; Chappell and Reid 2002). This finding also confirms previous studies in Thailand that positive aspects of the care-giving experience included the opportunity to show gratitude and accrue karma (from performing good deeds), a notion primarily rooted in Buddhist teachings (Gray et al. 2016; Pankong et al. 2018). Unexpectedly, religious coping using meditation or praying shows a negative association with SWB. A possible explanation is a reverse relationship (i.e. SWB predicts religious practices) or the lack of information on how frequently people engage in these Buddhist practices. The frequency of practice was found to have significant effect on SWB in an earlier Thai study (Winzer and Gray 2019). The acceptance of care-giving as a result of a bad deed done in the past is not significantly associated with SWB, but is in the expected positive direction.

With regard to control variables, as expected, caregivers who reported better health were happier than those with poorer health. This was also found in the study of Schulz and Sherwood (2008). The non-significant effects on SWB from gender and kinship with the care recipient may be due to the fact that specific gender roles in care-giving, from either the caregivers’ or the older recipient’s perspective, are not strong. Thai adults take on care-giving duties depending on both the task and the availability of relatives (Gray et al. 2016; Rittirong et al. 2014). The length of care-giving can reduce SWB (the wear-and-tear hypothesis (negative outcome)) or raise SWB [the adaptation hypothesis (positive outcome)] (Gaugler et al. 2005). The presence of these two forces may have led to a non-significant association of length of care-giving and SWB in this study.

This study has some limitations. In our analysis, SWB was used as the dependent variable. There is a concern that SWB can also be an independent variable (the issue of causality or temporal relationship) (George 2010). Longitudinal data will help explore this consequence of caregivers’ SWB. In addition, the data were from only one province in Thailand. Because our study is based on secondary data, our model does not rely on validated scales, but on items that were developed for the 2011 survey. However, most of the validated scales of PAC were used for caregivers who looked after older people outside of Southeast Asia with specific diseases. This study looks at older people with functional limitations in Thailand. For example, scales were developed for caregivers who were providing a) care for individuals with Alzheimer’s disease in the USA (Tarlow et al. 2004), b) care for patients with dementia in Hong Kong (Lou et al. 2015), and c) care for patients with acquired brain injuries in Spain (Las Hayas et al. 2014). There may be cross-cultural differences in questions, responses and measurement models. In addition, reliability tests are used for multiple-item scales, but most of our measures consist of a single-item.

Despite these limitations, this study has strengths and implications for policy makers and future research. Several studies outside Thailand as well as a Thai study support the focus on PAC in daily care-giving experiences and in the appraisal of gains in SWB while care-giving (e.g. Cohen et al. 2002; Pankong et al. 2018; Rapp and Chao 2000). This may be because PAC are important determinants of the quality of care provided to older persons (Kramer 1997). Caregivers typically provide better care when they perceive the care-giving experience to be satisfying and rewarding (Hodge and Sun 2012). On the other hand, caregivers’ stress can lead to elder abuse and neglect (Gupta and Chaudhuri 2008). This study suggests that an intervention to increase PAC as well as to decrease negative aspects of care-giving may provide better outcomes.

At the local level, to decrease the negative aspects and to increase the positive aspects of care-giving, resources in Thai communities should be mobilized to develop and to deliver intervention programs for older recipients and family caregivers. Organizations that are responsible for the well-being of older people in Thailand are available in all communities. These organizations include Buddhist temples (monks), senior citizens clubs (older people volunteers), sub-district health promotion hospitals (health personnel) and sub-district administrative organizations.

Despite the Thai policy that older persons should be supported by family members, there is very little formal social recognition of the family caregivers’ contributions to family and to society. The only available benefits for caregivers in Thailand are tax deductions for children who earn income as well as look after their parents or parents-in-law and for health insurance policies purchased by any child for their parents or parents-in-law (Ministry of Social Development and Human Security 2007). Thailand, as well as other Asian countries, should develop programs that recognize the contributions of family caregivers and provide support to them.

It has been suggested that the long-term care system in Thailand needs to be improved and further developed. As an Eastern country, Thailand has relied on limited public providers with imposed filial piety values and tax relief (Chuakhamfoo and Pannarunothai 2014). Lessons learned from a study in Thailand and Costa Rica, both middle income countries, show that there is some potential for community volunteers to contribute towards long-term care provision for older people in lower-resource countries. This should, however, be only one part of a broad group of interventions integrating health and social care. Bridging formal and informal providers of assistance to older adults should also be considered. Promoting long-term care volunteers is unlikely to be effective as a stand-alone policy (Lloyd-Sherlock et al. 2017). Additionally, lessons can be learned from high-resource countries in Europe, where the predominant formal care model has eroded over time. In recent years, due to demographic ageing trends and finite resources there has been a move from formal care towards integrated health and care systems that focus on care at home. Today, the vast majority of care for older people living at home is provided by family members. However, these countries still maintain a range of formal health and social care support services that are accepted by older adults and their family caregivers. (Erlingsson et al. 2012; Spasova et al. 2018).

Future research should examine the roles of both internal and external resources and caregiver well-being in other populations. Because findings (including potential explanations) and policy recommendations may be culture and area specific, studies of caregivers should be expanded to the national level as well as to other Asian countries where filial piety is a norm and family members are the main caregivers. This expansion of research may lead to a better understanding of the complexity of care-giving issues that emerge with the changes in demography, socioeconomic status and culture that occur over time.

Funding

The funding was provided by Office of the Higher Education Commission, Ministry of Education, Thailand, under the National Research University program (NRU).

Compliance with ethical standards

Conflict of interest

The authors declare no potential conflicts of interest with respect to the research authorship, and/or publication of this article.