Feeling Alone Together: Loneliness in Older Adults with Cognitive Impairment and Their Family Caregivers with Low Mastery

Abstract

Introduction

Loneliness in older persons with cognitive impairment (PCI) may beget loneliness in their family caregivers, depending on buffering resources caregivers possess. This study examined the association between loneliness in older PCI and loneliness experienced by their family caregivers, and the moderating role of caregiver mastery in this association.

Methods

Dyadic data from 135 PCI and their family caregivers in Singapore were analyzed using multivariable regression. Loneliness was measured using a three-item UCLA loneliness scale. Mastery was assessed using a seven-item Pearlin instrument.

Results

Multivariable regression showed that PCI loneliness and caregiver loneliness were weakly associated, taking other covariates into account. Notably, a significant interaction between PCI loneliness and caregiver mastery was observed, indicating that PCI loneliness was associated with caregiver loneliness only when caregivers had low mastery.

Conclusion

Lonely PCI may share their feelings of loneliness with their caregivers, and this can lead to loneliness among caregivers if they have low mastery. Promoting caregiver mastery may help reduce caregiver loneliness, directly and indirectly as a buffer against PCI loneliness.

Introduction

Loneliness, a discrepancy between the quantity and quality of social relationships one experiences and one desires [1], is a major public health concern worldwide [2, 3]. In developed countries, about one-third of people feel lonely and one in twelve suffer from severe loneliness, and the proportion continues to rise [4]. Loneliness takes a significant toll on physical and mental health. It is associated with higher rates of heart attack, stroke, depression, and anxiety [5, 6]. In turn, people with chronic loneliness have an increased risk of mortality, on par with smoking, obesity, and physical inactivity [6, 7].

Loneliness may be particularly prevalent among community-dwelling older persons with cognitive impairment (PCI) and their family caregivers [8, 9]. Cognitive decline may make it difficult for PCI to maintain the quantity and quality of formal or informal social engagement they once enjoyed, which may lead to loneliness. Family caregivers may also suffer from a lack of social contact due to the demands and responsibilities of caregiving. Some caregivers are socially connected but experience loneliness because the quality of their social relationships does not meet their expectations [4, 10]. However, we know little about whether and when loneliness among PCI and loneliness among their family caregivers are related. This study aimed to fill this gap.

The theory of emotional contagion [11] suggests that loneliness is transmissible [12]. Loneliness directly and indirectly affects one’s thoughts and behaviors [13]. In particular, lonely PCI tend to perceive their social surroundings as threatening, and express their worry, anxiety, and hostility accompanying their loneliness to their caregivers, with whom they spend most of their time [12]. Caregivers may capture and internalize loneliness expressed by their PCI and become lonely themselves. In short, loneliness in PCI and their caregivers can be associated because loneliness in PCI may beget loneliness in their caregivers.

However, not all caregivers of lonely PCI experience loneliness. The caregiver stress process model proposes that buffering resources in caregivers may mitigate the negative impact of caregiving stressors on caregiver health and well-being [14]. In particular, the buffering role of caregiver mastery on caregiver loneliness is well-established [15, 16]. Mastery, a sense of having control over one’s life [17], can reduce loneliness by altering the perception of loneliness and initiating counteracting behaviors [15, 16]. Therefore, caregivers with high mastery may be less affected by loneliness expressed by their PCI because they perceive their PCI’s loneliness as controllable and make behavioral changes to cope with the situation [18]. Overall, based on the emotional contagion theory and the caregiver stress process model, the present study aimed to examine the association between PCI loneliness and caregiver loneliness, and whether caregiver mastery moderates the association.

This study focused on loneliness among PCI and their caregivers in Singapore, a developed city-state in Southeast Asia. Previous research, conducted a decade ago, reported that approximately one in ten older adults living in the community in Singapore had dementia [19]. With the rapid aging of population [20], the prevalence of cognitive impairment among older Singaporeans is expected to rise. This necessitates the increasing role of family caregivers of PCI in the community. Given these circumstances, it is important to investigate whether and to what extent PCI loneliness contributes to caregiver loneliness. The insights gained from this study may inform the development of tailored policy initiatives aimed at addressing loneliness among caregivers.

Materials and Methods

Participants and Data Collection

We analyzed data on 135 PCI-caregiver dyads from the “Caring for persons with dementia and their caregivers in the community: Towards a sustainable community based dementia care system (COGNITION)” study [10]. The study received approval from the Institutional Review Board at the National University of Singapore. In 2018, 3,589 Singaporeans or permanent residents aged 60 years and above in the Whampoa community, which has a higher concentration of older adults in Singapore, were screened for cognitive impairment using the eight-item interview to Differentiate Aging and Dementia and two items (copying intersecting pentagon and three-item recall) from the Mini-Mental State Examination [21]. Of these, 323 individuals (9%) who scored 8 or less out of 10 were considered eligible for the study and 266 individuals (82%) and their caregivers consented to participate. A qualifying caregiver was a family member or a friend of PCI who was mostly involved in providing care or ensuring the provision of care to PCI. Both PCI and caregivers were interviewed face-to-face; however, as 131 PCI (49%) were unable to respond due to cognitive or health reasons, the caregiver responded as a proxy. This study focuses on PCI and caregivers’ self-reported experiences of loneliness. Therefore, the final analytic sample included 135 PCI-caregiver dyads (51%) in which the PCI answered the interview themselves (Fig. 1).

Fig. 1.

Analytical sample.

Measures

PCI and Caregiver Loneliness

Loneliness was measured using the 3-item loneliness scale, which has been widely used to measure perceived social isolation [22]. The scale includes three questions on the lack of companionship, feeling left out, and feeling isolated, with three possible responses: hardly ever = 1, some of the time = 2, and often = 3. The responses were added up to produce summated variables, PCI loneliness and caregiver loneliness, ranging from 3 to 9. Given the high number of zeros, we dichotomized the variable to distinguish caregivers or PCI who reported “some of time” or “often” to any item from those who answered “hardly ever” to all items [23].

Caregiver Mastery

Caregiver mastery evaluates perceptions of the extent to which one’s life chances are under one’s own control. It was assessed by a 7-item instrument, developed by Pearlin and Schooler [17]. This measure assessed a general sense of mastery in one’s overall life rather than the role-specific mastery attached to the caregiving role [24]. Respondents were asked to indicate the extent to which they agreed with the statement “there is really no way I can solve some of the problems I have” with four possible responses: strongly disagree = 1; disagree = 2; agree = 3; and strongly agree = 4. Negative items were reverse-coded and a summated variable was created, ranging from 7 to 28. A higher score indicated a higher level of mastery.

Covariates: PCI and Caregiver Characteristics

We accounted for other PCI and caregiver characteristics known to be associated with caregiver loneliness [9]. PCI characteristics comprised age (range: 61–98 years), gender (female = 1; male = 0), and memory/behavior/mood problems from the 24-item Revised Memory and Behavior Problems Checklist measured by caregivers, with a summated range of 0–96 [25].

Caregiver characteristics included age (range: 23–93 years), gender (female = 1; male = 0), minority ethnicity (Malay, Indian, and other nationalities = 1; Chinese = 0), married (married = 1; non-married = 0), highest completed education (no formal education = 1; primary school = 2; secondary = 3; post-secondary and tertiary = 4), social isolation (caregiver answered “hardly ever or never” to any of the four items: eating together with family; communicating with family or friends by phone or digital contact; having confiding relations; having someone to trust = 1, the rest = 0) [26], self-rated health (poor = 1; fair = 2; good = 3; very good = 4; excellent health = 5), caregiver burden (a 22-item instrument from the Zarit Burden Interview, ranging from 0 to 88) [27], adult child caregiver (caregiver is a son/daughter of PCI = 1; the rest = 0), and long-term caregiver (helping PCI with memory problems for more than 5 years = 1; the rest = 0).

Analytic Strategy

Since we dichotomized our outcome variable, caregiver loneliness, logit regression was employed. We first tested a direct association between PCI loneliness and caregiver mastery without covariates. Next, we added covariates to check the robustness of the association and then introduced an interaction term between PCI loneliness and caregiver mastery. The analytic models we tested showed a low risk of multicollinearity, with the moderate correlation between variables (online suppl. Table 1; for all online suppl. material, see https://doi.org/10.1159/000536644) and mean variance inflation factor less than two (online suppl. Table 2) [28]. We reported regression coefficients with 95% confidence intervals based on robust standard errors.

Results

Sample Characteristics

Table 1 provides sample characteristics. PCI had a mean age of 79 and were roughly evenly divided between males (48%) and females (52%). They scored on average 14 out of 96 for memory/behavior/mood problems, indicating relatively mild impairment as expected for PCI who were able to complete the interview without a proxy.

Table 1.

Descriptive statistics

Variable	Mean (SD)/%	Range
Dependent variable
Caregiver loneliness	3.49 (1.13)	3–9
Caregiver loneliness (dichotomized)	21.48%
Independent variable
PCI loneliness	3.74 (1.28)	3–9
PCI loneliness (dichotomized)	33.33%
Moderating variable
Caregiver mastery	12.90 (2.76)	7–28
Other covariates
PCI characteristics
Age	79.33 (8.55)	61–98
Female	51.85%
Memory/behavior/mood problems	14.33 (13.86)	0–96
Caregiver characteristics
Age	61.04 (14.38)	23–93
Female	58.52%
Ethnic minority	10.37%
Married	64.44%
Highest completed education	2.86 (0.95)	1–4
Social isolation	21.48%
Caregiver burden	22.90 (13.74)	0–86
Self-rated health	2.76 (0.80)	1–5
Adult child caregiver	50.37%
Long-term caregiver (>5 years)	5.93%

N = 135. SD, standard deviation; PCI, persons with cognitive impairment.

Caregivers had a mean age of 61, with the youngest caregiver being only 23 and the oldest being 93. This indicates that some caregivers were from the same generation as the PCI, while others were from following generations. Indeed, 50% of caregivers were adult children of the PCI. Caregivers were more likely to be female (59%), and 64% were married. One in ten caregivers was from an ethnic minority. On average, caregivers completed at least secondary education. About one in five caregivers reported “hardly ever or never” to any of the four social isolation items. Caregivers reported moderate levels of burden with a mean score of 23 on a range of 0–86. The average self-rated health score of caregivers was 2.76 out of 5, falling between fair and good. About 6% of caregivers had been caring for their PCI for more than 5 years.

Regarding the variables of interest, about one in five caregivers and one-third of PCI reported feeling “sometimes” or “often” left out, isolated, or lacking companionship. Caregivers rated their sense of mastery at a mean of 13 out of a possible range of 7–28.

Multivariable Regression

Table 2 presents the results of logit regression. In model 1, we found a weak association between PCI and caregiver loneliness. However, when caregiver mastery and PCI and caregiver characteristics were included in model 2, the association between PCI and caregiver loneliness was no longer statistically significant. After adding an interaction term between PCI loneliness and caregiver mastery in model 3, the association between PCI and caregiver loneliness became stronger and statistically significant. More importantly, we found the interaction term significant, indicating that caregiver mastery moderated the association between PCI loneliness and caregiver loneliness.

Table 2.

Association of caregiver social isolation profiles with caregiver burden: results from logit regression models

	Model 1		Model 2		Model 3
	β	95% CI	β	95% CI	β	95% CI
PCI loneliness	0.81+	[−0.03, 1.66]	0.48	[−0.51, 1.47]	6.18*	[0.93, 11.44]
Caregiver mastery			−0.19	[−0.46, 0.08]	−0.05	[−0.34, 0.25]
PCI loneliness x mastery					−0.48*	[−0.92, −0.04]
PCI characteristics
Age			−0.00	[−0.07, 0.06]	0.01	[−0.06, 0.07]
Female			−0.33	[−1.38, 0.73]	−0.36	[−1.36, 0.64]
Memory/behavior/mood problems			0.02	[−0.01, 0.06]	0.03	[−0.01, 0.06]
Caregiver characteristics
Age			0.02	[−0.03, 0.08]	0.02	[−0.03, 0.08]
Female			0.58	[−0.51, 1.66]	0.52	[−0.59, 1.62]
Ethnic minority			0.41	[−1.37, 2.19]	0.52	[−1.35, 2.39]
Married			−1.57**	[−2.62, −0.51]	−1.53**	[−2.62, −0.45]
Highest completed education			0.21	[−0.54, 0.96]	0.23	[−0.57, 1.02]
Social isolation			0.74	[−0.27, 1.75]	0.77	[−0.31, 1.84]
Caregiver burden			0.02	[−0.02, 0.06]	0.02	[−0.02, 0.06]
Self-rated health			−0.45	[−1.06, 0.17]	−0.55+	[−1.16, 0.07]
Children caregiver			0.19	[−1.64, 2.02]	0.32	[−1.46, 2.11]
Long-term caregiver			−0.50	[−1.94, 0.94]	−1.14	[−3.22, 0.94]

N = 135; CI, confidence interval; PCI, persons with cognitive impairment.

⁺ p < 0.10, *p < 0.05, **p < 0.01, ***p < 0.001.

To facilitate interpretation of the interaction terms, Figure 2 plots the estimated levels of caregiver loneliness by PCI loneliness at high (mean+1SD) and low (mean-1SD) levels of caregiver mastery. When caregivers had lower mastery, the likelihood of caregivers’ reporting loneliness tended to increase if their PCI experienced loneliness. However, such a relationship was not found when caregivers had high mastery.

Fig. 2.

Interaction between loneliness in persons with cognitive impairment and caregiver mastery in their association with caregiver loneliness.

Discussion

Loneliness has emerged as a global public health concern. However, less attention has been given to loneliness experienced by PCI and their caregivers. This study aimed to bridge this gap by investigating whether and how PCI and caregiver loneliness are associated, utilizing dyadic data from Singapore. Multivariable regression showed a weak association between PCI loneliness and caregiver loneliness, which lost significance after accounting for other covariates. Yet, the results need to be viewed in light of the moderating effect of caregiver mastery. Notably, the interaction between PCI loneliness and caregiver mastery was found to be significant, implying the buffering role of caregiver mastery in the association between PCI loneliness and caregiver loneliness.

The theory of emotional contagion [13] suggests that loneliness can be transmitted between individuals through social interactions, as lonely people tend to share their feelings of loneliness with others [12]. Our findings support this theory by showing an association between loneliness in PCI and their family caregivers. We further sought to extend the theory by identifying the moderating factors that condition emotional contagion. This is crucial because exposure to PCI loneliness may not yield uniform consequences due to variations in caregivers’ protective resources. Based on the caregiver stress process model [14], we found that PCI loneliness was significantly associated with caregiver loneliness only when caregivers had low mastery. The findings lend credence to studies on the stress-buffering role of mastery [29] and the mental health benefits of caregiver mastery among dementia caregivers [30–32].

However, similar to the reciprocal associations observed in loneliness among married men and women [33], it is also possible that caregiver loneliness could trigger PCI loneliness, with PCI mastery acting as a protective role against caregiver loneliness. To explore this possibility, we conducted a supplementary analysis considering caregiver loneliness as a predictor in its association with PCI loneliness as an outcome. Online supplementary Table 3 shows that caregiver loneliness had a limited association with PCI loneliness, taking other covariates into account. Also, no significant interaction was found between caregiver loneliness and PCI mastery in its association with PCI loneliness. Future studies may delve deeper into this reciprocal relationship using longitudinal data with advanced analytical models such as the actor-partner interdependence model [34].

The findings call for the need to design tailored interventions aimed at enhancing caregiver mastery. In particular, psychoeducational programs that empower caregivers by reevaluating the situation and developing coping strategies may bolster caregiver mastery [35, 36]. For example, Collins and Benedict [35] showed that community-based educational programs, delivered weekly over a 4-month period, increased mastery and reduced loneliness in older adults. Similarly, Boele et al. [36] found that psychoeducation and cognitive behavioral therapy over 8 months could improve caregiver mastery.

However, it is crucial to note that caregiver mastery tends to decline for caregivers who remain in the caregiving role for an extended period, compared to those whose PCI passes away or is admitted to an institution [37]. Therefore, it is also important to improve the availability and affordability of respite services, home-based and community-based care, and institutional long-term care to provide caregivers and PCI with sufficient alternatives [38]. In Asian societies, where institutional care is often surrounded by stigma due to obligatory norms of filial piety [39, 40], professional home-based and community-based services can relieve the burden on family caregivers [41]. In addition, peer group interventions have proven to be helpful in reducing caregiver loneliness [42], providing emotional support and psychosocial education that guide caregivers to better understand their role and regulate their emotions [42].

Despite the insights, this study had several limitations. First, its cross-sectional design raises concerns about endogeneity issues, including reverse causality and unaccounted confounders. In addition to the plausible reciprocal relationship between PCI and caregiver loneliness, situational factors such as financial adversity or stressful life events may bias the study findings. Second, the small sample size limited the subgroup analysis by gender, despite the fact that loneliness may spread more easily among women than men because women are more sensitive to others’ loneliness [12]. Third, the proportion of PCI and their caregivers experiencing loneliness in this study was relatively lower than that reported in Western contexts [43]. Although levels of loneliness among older adults tend to be lower in Asian societies [44, 45], future studies may revisit the validity of the instrument measuring loneliness in local languages. Fourth, the generalizability of our findings is constrained by its focus on a single community in a developed, small city-state before the COVID-19 pandemic.

Conclusions

Loneliness can be contagious because lonely people tend to share their loneliness with those who they frequently contact. This may be particularly true for PCI and their family caregivers, but also contingent on caregivers’ buffering resources. Using dyadic data from 135 PCI and their family caregivers, this study showed that PCI loneliness was associated with caregiver loneliness when caregivers had low mastery. This may be because caregivers with low buffering resources were more vulnerable to the spread of loneliness from their PCI. The findings contribute to the literature that has identified risk factors of caregiver loneliness. Practically, tailored interventions should be designed to promote caregiver mastery to protect against caregiver loneliness when their PCI experience loneliness.

Statement of Ethics

The caring for persons with dementia and their caregivers in the community: Towards a sustainable community based dementia care system (COGNITION) study received approval from the Institutional Review Board at the National University of Singapore (approval number H-17-013). Written informed consent was obtained from both persons with cognitive impairment (PCI) and their caregivers in the COGNITION study prior to survey administration. If PCI were unable to respond due to health reasons, consent was obtained from PCI’s legal representative or next of kin.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

Caring for persons with dementia and their caregivers in the community: Towards a sustainable community based dementia care system (COGNITION) study was supported by the National Innovation Challenge on Active and Confident Ageing Grant (award no.: MOH/NIC/COG05/2017). The funding organizations had no role in the design or conduct of the study; collection, management, analysis, or interpretation of the results; or preparation, or approval of the manuscript.

Author Contributions

Pildoo Sung designed the study, conducted data analysis, and wrote the manuscript. Jeremy Lim-Soh contributed to the writing and critically reviewed the manuscript. Angelique Chan, the principal investigator of the COGNITION study, managed the project and collected the data.

Funding Statement

Caring for persons with dementia and their caregivers in the community: Towards a sustainable community based dementia care system (COGNITION) study was supported by the National Innovation Challenge on Active and Confident Ageing Grant (award no.: MOH/NIC/COG05/2017). The funding organizations had no role in the design or conduct of the study; collection, management, analysis, or interpretation of the results; or preparation, or approval of the manuscript.

Data Availability Statement

The complete COGNITION study dataset is not publicly available due to legal and ethical reasons. The data used in this study are available upon reasonable request. Further inquiries can be directed to the corresponding author.