Psychosocial factors and caregiver burden among primary family caregivers of frail older adults with multimorbidity

Provision of care for frail older adults with multiple chronic diseases (multimorbidity) poses increasing challenge for family caregivers. Frailty, in addition to multimorbidity, can result in increased family caregiver burden especially when increased time spent caregiving was necessitated [1]. Several studies have shown evidence of caregiver burden especially in caring for older adults with increasing dependency [2‐4]. For older adults with frailty, the loss in body reserves as a result of exposure to stressors increases vulnerability to adverse health outcomes [5]. For the caregivers, stress can arise from the associated financial responsibilities, role changes, assisting in activities of daily living (ADL) and dealing with potential behavourial problems in caring for the frail care recipients [6]. Caregiver stress is a consequence of several interrelated events and directly related to hardship and problems experienced in caregiving and role strain outside of caregiving [7]. Long term demands for caregiving can result in decline in psychological health, anxiety [8], maladaptive coping [9], decreased care provision and decreased quality of life for both caregivers and care recipients [10].

Competency refers to the skills needed by an individual to perform a task at a certain level of performance [11]. A study showed that high competence had protective effect against psychological burden of caregiving for caregivers of persons with dementia and suggested there may be similar positive association for caregivers of frail older adults [8]. However, previous care experience was not associated with improved caregiving effectiveness if the complexity of care needs of a dependent person rises. This in turn, results in a higher level of burden and sense of insecurity [12].

Positive aspects of caregiving (PAC) has mediating effects on caregiver burden [13]. PAC optimizes positive experience and reduces stress as it promotes a sense of personal accomplishment and gratification, increases family cohesion, sense of personal growth and purpose in life [14].

Social support is a multi-dimensional construct which comprises instrumental support, such as help with daily tasks; informational support and emotional support: such as having someone to discuss problems [15]. Social support may refer to actual received specific support, or perceived, that is, the subjective appraisal of adequacy and quality of support. A meta-analysis found that perceived social support has a greater effect on caregiver burden than received support [16].

Variations of the definition of multimorbidity existed in different literature with most literature citing as the coexistence of two or more chronic conditions [17]. Our study adopted the cut- off of multimorbidity as three or more chronic medical conditions based on a systematic review [18]. Using three or more chronic medical conditions enabled more discrimination in identifying patients with higher needs, and was found to be more clinically useful and specific to older adults [19]. In Singapore, local prevalence of multimorbidity was found to be higher in patients with increasing age, male gender, and those of Chinese and Indian ethnicity [20]. As compared to a non- Chinese primary caregiver, a Chinese primary caregiver had a 3 times odds of perceiving burden [1].

Older adults with both frailty and multimorbidity may increase in caregiver burden but presence of perceived caregiving competence, positive aspects in caregiving and social support may alleviate this burden. There has been limited research on caregiver burden of caregivers of frail, multi-morbid older adults, particularly in Asian context. The impact of psychosocial factors on caregiver burden for this population has not been explored. Hence, we conducted this study based on the original research from Ding et al. (2022) [1] study, and further explored the psychosocial aspects of family caregiver burden using the existing data.

This was a descriptive cross-sectional study utilizing interviewer-administered questionnaires.

The ‘rule of thumb’ states that a minimum of 10 participants per predictor variable is needed to determine sample size for regression equations [21]. Hence, for 18 independent variables, a minimum of 180 participants was needed. Participants were recruited from two public primary care clinics (Polyclinics) in Singapore over 5 months from July 2020 to December 2020. Polyclinics served to provide a one-stop health service centre providing outpatient medical treatment, health promotion and chronic disease management. Consecutive recruitment was done for caregivers who fulfilled the criteria of informal caregiving for at least three months, (i.e., they did not receive a salary for caregiving), were family members of the care recipients, were primary caregivers most involved in providing or ensuring care for care recipients and were able to speak English or other local languages. Their care recipients were aged 65 and above, not institutionalized, had multimorbidity with at least three chronic medical conditions from Fortin et al., 2017 [22] list of conditions for primary care, with a Clinical Frailty Scale (CFS) [23] score 4 and above, indicating mild to severe frailty. Those who were receiving cancer treatment or palliative care were excluded as their specific caregiving needs were not within the purview of primary care.

The study was approved by the National Healthcare Group- Domain Specific Review Board (NHG DSRB Ref: 2020/00014). Verbal informed consent was obtained from all participants.

Questionnaires were administered by trained study team members proficient in the participants’ spoken language in the primary care clinics. Data collected were entered by the study team electronically via the Research Electronic Data Capture (REDCap ®), an intranet accessible platform [1]. REDCap ® is a web- based software developed by Vanderbilt University that capture research data in a systematic manner [24].

The questionnaire comprised of caregivers’ sociodemographic variables such as age, gender, education, working status, presence of alternative caregivers and use of community resources. The perceived caregiving competence, positive aspects of caregiving and social support were measured using a validated questionnaire- Caregiving Competence Scale (CCS) [8], Short Positive Aspects of Caregiving (S-PAC) [25] and the modified Medical Outcomes Study Social support survey (mMOS-SS) [23] respectively. Caregiver burden was measured using the 12-item Zarit Burden Interview (ZBI) [27].

The CCS is a four-item scale that measures self-appraisal of one’s efficacy in caregiving. Scoring ranged from not at all (1) to very much (4). A possible score ranged from 4 to 16 with higher scores corresponding to higher levels of competency in caregiving. The CCS has Cronbach’s alpha of 0.74 suggesting it has good internal consistency. It was used locally for caregivers of frail persons in a hospital context [8]. A Chinese version of CCS has been developed and psychometric properties evaluated on family caregivers of stroke survivors. It also demonstrated good internal consistency (Cronbach’s alpha of 0.81) [28].

The S- PAC consists of seven items, which was validated among a more general caregiver population comprising caregivers of home dwelling older adults with functional limitations [25] and frailty [29]. Five items of the self-affirmation (SA) subscale and two items on the outlook-on-life (OL) subscale are scored on a five-point Likert scale ranging from disagree a lot (1) to agree a lot (5). A higher score indicates a more positive perception of caregiving experience. The S-PAC demonstrated good internal consistency in our local setting with a Cronbach’s alpha of 0.91 (overall scale) [25].

The 8- item mMOS-SS consists of two subscales covering two domains (emotional and instrumental [tangible] social support) with four items each. Participants were asked on how often social support such as companionship or assistance were available to them when they needed it. It was measured on a 5- points Likert scale ranging from none of the time (1) to all of the time (5). Scores were calculated as the average score of subscale items transformed to a zero to 100 scale with higher scores indicating more support. The mMOS-SS was a valid and reliable measure of social support especially for geriatric assessments [26]. Across populations, the internal consistency of the mMOS-SS measure was very good (Cronbach’s alpha 0.88 to 0.93).

The 12-item ZBI measured subjective caregiver burden. It scores ranged from 0 to 48, with higher score indicating greater burden. Each item was scored on a 5- point Likert scale from 0 to 4 (never to nearly always). For this study, a cut off score of 10 or above was used to determine that family caregivers perceived caregiver burden. This cut-off was based on the study of caregivers of frail older persons by Mello and team [30] as this was most similar to our population of interest. The 12-item ZBI was validated in other caregiver populations such as informal caregivers of elderly persons irrespective of level of cognition (Cronbach’s alpha 0.90) [27], caregivers of community-dwelling older adults with diverse comorbidities (Cronbach’s alpha 0.81) [31] and caregivers of frail older persons [30]. The 12-item ZBI was also used and validated locally, but for caregivers of persons with dementia [32].

A pilot study of 20 participants was conducted to determine face validity and test–retest reliability of the overall questionnaire in local context. The questionnaire was understandable and no amendments were needed. The intraclass correlation (ICC) for test–retest reliability was 0.932 (CI 0.919, 0.937, p < 0.001) indicating excellent reliability [33].

A total of 205 eligible caregivers were approached, of whom 188 agreed to participate, giving a response rate of 91.7%. The characteristics of the caregivers and total mean score for CCS, S-PAC, mMOS-SS and ZBI score are shown in Table 1.

Table 2 presents the individual item scores for perceived caregiving competence, positive aspects in caregiving and social support of the caregivers for all participants. Most of the respondents agreed that caregiving “made them feel useful”, “important” and “appreciated”. In terms of social support, more than one-third of caregivers had someone to turn to for suggestions most of the time.

Mann–Whitney U test was used to compare differences in perceived caregiving competency, positive aspects of caregiving and social support between caregivers who perceived burden versus those who did not perceive burden. As shown in Table 3, caregivers who perceived burden had significantly lower perceived caregiving competency, significantly lower perceived positive aspects of caregiving and significantly lower perceived social support.

Multivariable logistic regression was used to analyse factors associated with caregiver burden (Table 4). Factors significantly associated with higher odds of perceived burden were living in 4 or 5 room housing development board (HDB) or executive flats (OR 4.4, 95% CI 1.39, 13.72, p = 0.01), presence of alternative caregivers (OR 3.3, 95% CI 1.09, 10.19, p = 0.04), use of community resources (OR 4.4, 95% CI 1.15, 16.83, p = 0.03) and time spent caregiving per week (OR 1.1, 95% CI 1.02, 1.10, p = 0.003). Higher caregiving competency (OR 0.7, 95% CI 0.58, 0.96, p = 0.02) and social support (OR 0.8, 95% CI 0.74, 0.89, p < 0.001) were significantly associated with lower odds of perceived burden. No statistically significant differences were found between caregiver ZBI score and age, gender, marital status, educational level and employment status of caregiver. Also, no difference was found in terms of relationship with (child vs spouse), or living with care recipient, or duration of caregiving with ZBI score.

The aim of our study was to evaluate to what extent psychosocial factors can mitigate the effect of burden experienced by family caregivers of frail older adults with multimorbidity. In the present study of 188 caregivers, we found that caregivers who perceived burden had significantly lower perceived caregiving competence, positive aspects of caregiving and social support scores. In addition, presence of alternative caregivers, community resources and longer time spent in caregiving were associated with caregiver burden.

Our study illustrated that psychosocial factors (perceived caregiving competence and social support) influenced caregiver burden among older adults with frailty and multimorbidity. These findings enhanced our understanding of caregiver burden with increasing care dependence (frailty) of our elderly population in primary care setting. This highlights the need for clinicians to assess and address psychosocial factors associated with caregiver burden. These include competence needs of both primary and alternative caregivers as well as caregivers’ awareness of community resources and formal support services. Policy makers should incorporate caregivers’ opinions to enhance resources for caregivers which are tailored to their needs. The findings added to evidence about the significance of perceived social support as a predictor of caregiver burden. The novel findings that alternative caregivers and the use of community resources may increase the likelihood of caregiver burden necessitate further research in this area. Whilst this study did not confirm the significant positive aspect of caregiving association with caregiver burden, more than half the caregivers reported positive feelings and felt being useful, appreciated and important when providing care. Future research could consider longitudinal study to evaluate the effects of psychosocial factors over time. Research could also evaluate interventions to improve PAC, caregiving competence and social support. It will also be useful to determine which aspects of caregiving are perceived to cause most burden, as addressing the root cause may further alleviate burden.