Motivations for being informal carers of people living with dementia: a systematic review of qualitative literature

Background Informal, often family carers play a vital role in supporting people living with dementia in the community. With ageing populations, the part played by these carers is increasing making it important that we understand what motivates them to take on the role. This systematic review aimed to identify and synthesise qualitative literature describing what motivates people to care for someone with dementia. Methods The review followed the Centre for Reviews and Dissemination (CRD) guidelines. Six electronic databases were searched from their first records until August 2018. Synthesis was narrative. Results Twenty-six studies fitting the inclusion criteria were identified. Carers described multiple, inter-related motives for caring for someone with dementia. Caring was generally described as a reflection of long-standing family relationships between carers and the care recipients, whether by blood or marriage. Commonly offered motivations included love, reciprocity, filial piety, duty and obligation. Conclusions Perhaps the most striking finding was the similarity in these motivations irrespective of gender or relationship with the care recipient. Family relationship and shared history underlay most motivations. Future research should include more longitudinal studies incorporating within study comparisons between different demographic groups to give greater confidence in identifying similarities and differences between demographic groups.


Background
Across the world, carers play a vital part in caring for their ill and disabled family members and friends. These informal, unpaid, often family carers (or caregivers as they are also known) are commonplace. For example, in the United Kingdom (UK) there are approximately 6.5 million carersone in eight adults [1]. In the United States of America (USA), the proportion is slightly higher where approximately 18% of the population are carers [2].
Being an unpaid carer is recognised as a mixture of satisfactions [3,4] and challenges, often with negative effects on carers' physical and mental health and can result in financial hardship for carers [5]. Carers of people living with dementia are particularly adversely affected by their role and report more mental health problems and stress than other carers [6]. Evidence also shows that the symptoms of dementia in care recipients are strongly associated with carer depression, stress and lower quality of life [7]. Importantly, unlike some caring roles, dementia is a progressive disease with an inevitable decline and although the duration of the caring role varies considerably, it sometimes lasts over a decade [8].
As populations worldwide are ageing, numbers of people living with dementia are rising. For example, it is predicted that the number of people with dementia in the UK will rise from approximately 850,000 to two million by 2050 [9]. Worldwide evidence suggests that most people with dementia live in the community and their primary support is from their families. For example, in the UK, two-thirds of people with dementia live at home, receiving most of their support from family carers [10].
Given the often challenging nature of the caring role, it is important to understand what initially motivates people to become carers of someone with dementia but it is also important to understand why they continue, despite the often growing needs of the care recipient.
Past research has highlighted the significance of carers' motivations for caring both for carers themselves and those they care for. A review of quantitative literature [11] looked at the impact of carer motivations and 'meaning' on the wellbeing of carers of people with dementia. Motivations for caring were described as why carers take on the role, whilst meaning related to how positive the experience of caring was for the carer. Ten studies were identified, six related to meaning and four to motivations. Their synthesis suggested that carers' kin-relationship to the care recipient and cultural norms influenced motivations to care and that carer wellbeing could be influenced by these motivations. Finding meaning had a positive impact on carer wellbeing. They also reported that carers' cultural background influenced motivations for caring. Some cultural groups emphasised the role played by religion whilst others highlighted filial responsibilities more than other groups. Overall, the evidence suggested that carers from Western cultures were less likely to refer to filial piety than carers from other cultures. The authors concluded that motivations and meanings can overlap conceptually and that more qualitative investigations are needed in order to understand caring motivations [11].
Importantly, quantitative research has demonstrated that the explanations non-spousal carers offer for caring are associated with care recipient outcomes, relationship quality and carer wellbeing. For example, being unwilling to take on the caring role was associated with higher abusive behaviour towards care recipients [12] and a quantitative questionnaire study found significant relationships between intrinsic motivations, finding meaning in caring and current relationship quality [13]. In another study, family carers of people with dementia who described more external pressure as caring motivators, as opposed to giving personal reasons, also reported more emotional health difficulties [14]. Similarly, Livingstone et al. [15] found carers' emotional health had an impact on care breakdown, institutionalisation and elder abuse.
Evidence such as this highlights the centrality of carers' motivations. However, we were unable to identify any recent reviews or syntheses of the qualitative literature. It was therefore decided to focus on qualitative literature to permit synthesis of in-depth research that describes carers' motivations in their own words and from their perspectives.

To identify and describe informal carers' motivations
for caring for people living with dementia, including their motivations at the start of caring and motivations for continuing to care. 2. Where possible to qualitatively identify and describe any similarities or differences in motivations amongst different demographic groups e.g. in terms of gender and relationships (e.g. spouse versus adult child) and ethnic or cultural groups.

Methods
The review followed the Centre of Reviews and Dissemination (CRD) guidelines [16] and was reported using the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines [17]. We included both the terms 'meaning' and motivations' in the searches as suggested by Quinn et al. [11]. The terms filial piety and filial obligation are key words in the Quinn et al. paper and so were also included. Other included search terms (e.g. loyalty, love, caring intention and best placed) were found through a combination of thesaurus searching for similar terms and searching of related MeSH terms in Ovid Medline.

Study screening and selection
Following duplicate removal, both review authors independently screened the titles and abstracts to identify studies potentially fitting the inclusion criteria. Both authors then scrutinised full texts of the selected articles.
Where there was uncertainty about inclusion, consensus was achieved by discussion.

Data extraction and management
Data were entered into standardised tables which included study aims, methods, findings and authors' conclusions.

Data synthesis
The research questions were broad, therefore narrative synthesis was selected [18]. Data sources were generally similar with most findings coming from semi-structured interviews thereby reducing some of the potential challenges of multi-method reviews. This synthesis process involved both the review authors independently reading, identifying, recording and summarising the themes relating to caring motivations as described by the original study authors. After this preliminary synthesis, the authors then met and discussed the themes identified and whether, if any, there were identifiable relationships, between the study findings and for example, the participants or geographical locations. In the light of these conversations and the identified motivations, the authors then revisited the included papers again. The very few minor differences between the authors' analysis after this process were resolved by discussion.

Quality appraisal
The quality of included studies was assessed independently by both authors using a modified version of a qualitative research rating scale [19]. This scale includes assessment of, for example, whether the methods were appropriate and whether there was clear connection to an existing body of knowledge or wider theoretical framework.
Studies were not excluded based on quality scores but this assessment process allowed interrogation of study methodological quality highlighting strengths and weaknesses.

Results
Electronic searches identified a total of 3482 articles before duplicate removal: MEDLINE -838; Embase -1241; PsychINFO -702; CINAHL Plus -168; ASSIA -282; and Scopus -251. After duplicate removal, the titles and abstracts of the remaining 2586 articles were scrutinised and 67 full texts were retrieved. From the searches, four relevant systematic reviews [11,[20][21][22] were retrieved for reference list checking. This process and articles suggested by included study authors revealed 10 additional studies for full text scrutiny.
In total 26 studies fitted the inclusion criteria ( Fig. 1).

Studies excluded but which came close to inclusion
Seven studies scrutinised at the full-text stage came close to inclusion but were eventually excluded for the following reasons. Harris [23] and Mukadam et al. [24] provided insufficient relevant data for inclusion. Harris et al. [25] was excluded as less than half the care recipients had dementia. Miyawaki [26] described filial responsibility in carers of frail elderly relatives but the proportion with dementia was unclear. With both Cahill [27] and Kabitsi and Powers [28], it was not possible to separate qualitative from quantitative findings. Caldwell et al. [29] focused primarily on carers' motivations for care home placements and not for continuing to care.
To help answer the review questions, the synthesised findings are firstly presented overall and then by carer demographic characteristics (relationship, gender, ethnicity, nationality, culture and religion). Where studies included a variety of relationships with the care recipients, motivations were often described for participants together and it was not always possible to be confident whether the motivations related to specific relationship       Table 4.

Overall
Carers described multiple, inter-related motives for caring for someone with dementia. Motivations did not appear to have any association with study publication date or study quality. Caring was generally described as a reflection of long-standing family relationships between carers and care recipients whether by blood or marriage. Commonly offered motivations included love, reciprocity, filial piety, duty and obligation. Some carers emphasised caring through choice whilst others described feeling obliged to take on the role whether this was from societal, cultural or family pressures or because available formal care was perceived to be poor quality or expensive. Some carers also said they were in the caring role to avoid care home admission [32,36,41,45,48,55]. Avoidance of guilt was also a motivator [30,41,47].

Ethnicity, nationality, culture and religion
Although there were some differences, motivations for caring were largely similar across ethnic and cultural groups. Many studies did not specify the ethnic or cultural groups of their carer participants, therefore in the following section, where ethnic or cultural groups were not supplied, review findings are reported by study country.
Reciprocity, love, duty, marital commitment and responsibility were frequently highlighted across carers in all included studies. However, the satisfactions derived from caring were more often highlighted by carers from Western countries [39,40,43] but it was also mentioned by Pang and Lee [44]. The term filial piety was more frequently used by authors describing the motivations of carers from Far Eastern groups such as Japanese [36,55], Chinese [32], Korean [38] and Vietnamese [43]. However, other authors referred to e.g. filial obligation [35,37]. Many explicitly emphasised the importance of family members caring for each other in their cultures as a reason for caring [33,34,37,38,49,51,53].
With the exception of three studies [36,52,54], few made clear within study comparisons between ethnic and cultural groups. Those that did suggest that, although there are some differences between groups, generally there are more similarities than differences.
Avoidance of nursing home admission was mentioned in studies that included carers from several ethnic and cultural groups including Chinese [32], Vietnamese [41], British [47] and Dutch [52].
Qadir et al. [46] reported that duty and obligation were influenced by the carers' Islamic faith. However, although religion and caring motivations were highlighted in several studies from several countries, the precise religion was not always specified [31,44,49,52,55].

Gender
Many of the motivations to care were similar for men and women and included reciprocity, love, responsibility and duty. However, caring was sometimes explicitly linked to being wives, daughters or daughters-in-law with some females describing caring as part of general female caring role within the family [51,54] or part of  This study underlines the importance of not only identifying carer physical & psychosocial features but also existential ones. Staff need to be more aware of these issues in order to support families also in existential crisis.
Cahill (2000)    their female identity [33]. For Japanese women, being a carer was a 'matter of course' (p69, [54]). Three studies offered insight into son carers [35,36,40]. For Japanese sons, birth order was important with the eldest typically responsible for providing care for relatives [36]. Findings from the other two studies suggested that sons, like other carers, were largely motivated by love and duty [35,40].

Continuing to care
It was not always possible to separate reasons for continuing to care from more general motivations but where it was possible, findings suggested a range of motivations for continuing to care. Some of these were the same as more general reasons for caring, for example, love and duty and part of their marital relationship [36,38]. Other reasons included avoiding paying for nursing homes [55], distrust of nursing home care or regarding their care as superior to institutional care [32,52]. The belief that carers could provide better care than professionals as they would care with more love and would provide care more tailored to individual needs [52] was also highlighted. Continuing companionship and satisfaction [43,44] and the belief that no one else was available [47] or that they would regret it if they did not continue in the role [55] were also highlighted.

Discussion
This review brought together studies from a wide range of cultures and countries covering nearly three decades of research. Many recent studies published since 2010 were identified suggesting a growth in interest in the topic. There were usually multiple explanations for caring but perhaps the most striking finding was the apparent similarity in carers' motivations for caring irrespective of their relationship with the care recipient, country of origin, ethnic or cultural background or gender. Motivations highlighted were frequently based on long-standing family relationships and emotions such as love and wanting to reciprocate or return care received from the person with dementia in the past. Caring was not generally regarded as something extraordinary but rather as a natural part of family life [36] and something 'you just get on with' (p229, [47]). Importantly, carers had multiple motivations for caring and it was rarely driven by one motivation alone. This is perhaps not surprising. Caring for someone with dementia is a fluid, complex, often long-term role and it is likely that the motivations and the mixture or balance of motivations change as the health of the person with dementia, the carers' own health and their situations change. There were only four longitudinal studies [33,39,42,54] and these tended not to emphasise any changes over time. However, Lin et al. [39] described how carers learnt from their caring experiences whilst responsibility and duty increased over time but support from formal and informal sources fluctuated.
Our review and the one by Quinn et al. [11] have shown that cultural factors play a role in motivations for being a carer but importantly emphasis on cultural differences may be misleading. Throughout these studies caring was underlain by long-standing relationships, often between close family members irrespective of carers' cultural background. This is a significant finding as it suggests motivations for caring may not be so very different across cultures as much of the literature appears to suggest. Caring was usually underwritten by love and respect for the person with dementia. Although some investigations did make specific within study ethnic or cultural comparisons, these were in the minority and they reported more similarities than differences in motivations [36,52,54].
Caring was frequently described as duty, obligation or responsibility. Filial piety or filial obligation was often explicitly given as an explanation for caring in studies including carers originally from China [32,37] Korea [38], Vietnam [41] and Japan [54,55]. Filial piety can be defined as a principle in Chinese and other Asian cultures, which 'emphasizes honour and devotion to one's parents… implies that adult children have a responsibility to sacrifice individual physical, financial and social interests for the benefit of their parents or family' (p14, [22]). However, carers from a wide range of other countries including South Africa [34], the USA [35], the UK [39], Pakistan [46] Brazil [49] and the Netherlands [52] also referred to caring for very similar reasons to those relating to filial piety, such as duty, obligation and responsibility. Again, this emphasises the commonalities across cultures.

Limitations of the included studies
The quality of the studies was variable with many omitting to report important methodological and participant demographic details. However, we did not detect any associations between study quality and reported findings. Synthesis of the studies was hampered by the fact that some studies did not report details such as carer ethnicity and when they did, they did not always present them in the same way. This meant that we often had to rely on the country of data collection and we were sometimes unable to comment on the relationship between ethnicity and motivations. Future research should ensure such details are reported.
Study findings were not presented in a manner that permitted identification of what, if any, motivations were seen as the primary or most important motivations by carers and although, perhaps tempting, it is important not to assume that the most frequently mentioned motivations were the most significant and carried the most weight for carers. The carers here may have responded with motivations that they perceived were most socially acceptable and the motivations offered may not have been the most important reasons for caring for them.
Indeed, carers may also not often consciously consider what motivates them to care for others, perhaps just regarding it as an unquestioned part of family relationships [19]. This may make it harder for them to articulate their motivations in research interviews and as a result, they may have relied on socially desirable explanations such as 'love' or 'duty'. Furthermore, social desirability may have made it difficult to admit to not wanting to care or choosing to care to save money on care home costs.
Any study investigating carers' experiences is hampered by the fact that many people that researchers and service providers regard as carers may not see themselves as carers. For them, caring may be seen as a progression in their relationship [56] and they may make few, if any, clear distinctions between spousal or kinship support as opposed to carer support [19]. They therefore may not volunteer to participate in research and as a result their experiences may not be reported. This important group of unidentified carers requires greater investigation in order to capture the motivations of all carers.

Review strengths and limitations
Our review was comprehensive and wide reaching and included many diverse studies. We believe the decision to include both 'motivation' and 'meaning' as search terms was justified given the conceptual overlap in the two concepts [11]. The decision to exclude quantitative studies to allow carer participants to describe their motivations in their own words, excluded potentially relevant studies. Despite this, 26 studies were included in the review.
There were some challenges to undertaking the review. For instance, although most studies only collected data using single one-to-one interviews, variability in reporting the data made synthesis more challenging. We also attempted to further understanding of similarities and differences in caring motivations by broad demographic groups, but clearly these distinctions are compounded because of the overlap between, for example, being female and also being a wife.
Caring is often long-term and is a changing mixture of challenges and rewards but most of the studies here were cross-sectional making it impossible to be confident about any changes in motivations for caring over time. Importantly this meant we were unable to answer confidently one of our review questionswhether motivations for starting to care for someone with dementia were similar or different to motivations for continuing to care. It may be that carers themselves do not make this distinction in their motivations but future research could perhaps focus specifically on this question to help understand why they continue to care even in adverse situations.

Conclusions
Future research should include more longitudinal studies and should make within study comparisons between different demographic groups (e.g. ethnic and gender groups) in order give greater confidence in identifying similarities and differences between demographic groups.
Articulating motivations for caring in these studies may also imply some sort of conscious choice by carers but for many caring may not be perceived as a choice but perhaps something that happens gradually in the context of mutual caring and support, especially amongst older spouses. Understanding the development of this, often mutual, caring deserves more research.
It is perhaps tempting to focus on dissimilarities between demographic groupswhether these relate to differences in gender or cultural or ethnic background. However, it is equally important to highlight commonalities amongst these groups. The studies included in this review show not only the diversity in motivations for providing care in diverse cultural settings but also the importance of families and the common threads of compassion and love for care recipients. For the participants here, their relationships and their affective bonds predated their caring role [57], emphasising again the similarities across cultures.
Participants here were all already in the caring role but it is also very important to improve our understanding of why people feel unable take on the caring role. The readiness of others in the family and the availability, costs, quality and suitability of residential care and other formal care are obvious factors and are similar to why carers decide whether to move the care recipient into residential care [58]. Future research should investigate this.
Abbreviations ASSIA: Applied Social Sciences Index and Abstracts; CINAHL Plus: Cumulative Index to Nursing and Allied Health Literature; CRD: Centre of Reviews and Dissemination; MeSH: Medical Subject Heading; PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analysis; UK: United Kingdom; USA: United States of America