Quality of life among older informal caregivers in Sweden: the role of loneliness and social isolation

Older adults are increasingly taking on caregiving roles due to gaps in formal care services. This study examines how loneliness and social isolation correlate with the quality of life (QoL) among older caregivers, identifying distinct profiles within this group. Using data from 994 caregivers aged 60 and older in the Swedish National study on Aging and Care in Kungsholmen (SNAC-K), researchers classified caregivers into three QoL profiles: good (57.9%), moderate (34.8%), and moderate physical, poor mental QoL (7.3%).

Caregivers in the poorest mental health profile were predominantly female spouses providing the most hours of care. Loneliness and social isolation were strongly associated with being in this group compared to those with good QoL. Additionally, men with higher levels of social isolation were more likely than women with similarly high levels of isolation to experience worse QoL, suggesting a gendered impact of social isolation on QoL.

These findings reveal that many older caregivers face significant mental and social challenges, particularly those providing intensive care or experiencing high isolation. The study emphasizes the urgent need for targeted policies to alleviate caregiver burden and improve their QoL through support programs addressing mental health, loneliness, and social connections.

The demand for medical and social care is rising due to aging populations worldwide, driven by declining fertility rates [1] and increasing life expectancy, especially among the oldest old [2]. In Sweden, while the welfare system provides substantial support through home care and residential care for people in need, these services are becoming increasingly insufficient [3‐5]. As a result, informal caregivers are taking on a significant share of the caregiving activities, often stepping in where formal care falls short.

In light of the increased prevalence of informal caregiving, its impact on caregivers’ lives has been a topic of interest. While some research suggests that the physical and mental demands of caregiving can negatively impact caregivers’ well-being, leading to heightened stress, anxiety, and physical illness [6‐9], other studies have identified potential positive outcomes associated with caregiving, such as a sense of purpose and fulfillment [10‐12]. Recently, increased attention has been devoted to social isolation (objective lack of social contacts) and loneliness (subjective feelings of being lonely) among caregivers. Most studies report increased loneliness, particularly among spousal caregivers [13, 14], but some suggest that caregiving may reduce caregivers’ loneliness by fostering meaningful connections [15]. In addition, previous studies have shown that loneliness and social isolation are not highly correlated (r = 0.201 [16]), and predict health outcomes through distinct pathways [17], warranting their consideration as separate, albeit connected, entities.

There is limited evidence on the impact of caregiving on older informal caregivers, who often face their own age-related health challenges and poorer well-being. Understanding their specific needs is thus essential for developing tailored support services. One way well-being has been operationalized is through quality of life (QoL), a concept that encompasses both physical and mental aspects of well-being. QoL measures are known to predict health outcomes, healthcare utilization, and the effectiveness of interventions [18].

This study aims to: (1) identify and describe profiles of older informal caregivers (60 years and above) in Sweden based on their QoL, and (2) explore the relationships between loneliness, social isolation and the different QoL profiles.

The primary outcome assessed was QoL, measured using the items from the SF-12 instrument modified for the SNAC-K population study, which includes 11 of the 12 original items (missing item 7: “Did work or activities less carefully than usual”). SF-12 is a widely used tool, comprising six physical health-related (items 1, 2, 3, 4, 5, 8, amounting to 20 points) and six mental-health related (items 6, 7, 9, 10, 11, 12, amounting to 25 points) questions [19]. We used 11 items available in SNAC-K and derived distinct profiles of caregivers based on their responses, using data-driven techniques described below. To describe caregiver QoL profiles, sociodemographic characteristics such as sex, age, education level (elementary, high school, university), occupation (manual, non-manual worker), and marital status (unmarried, married, divorced, widow) were used. To approximate socioeconomic position, education was included instead of income because it is a stable indicator across the life course and remains meaningful after retirement [20]. Lifestyle factors such as smoking (never, former, current), alcohol consumption (never/occasionally, light-moderate, heavy), body mass index (underweight, normal weight, overweight, obese), and physical activity level (inadequate, health-enhancing, fitness-enhancing) were also considered using previous operationalizations in SNAC-K [21]. Health status was assessed through three indicators. Namely, the Health Assessment Tool [22], which takes values from 0 to 10 (higher scores indicating better health), the Montgomery-Åsberg Depression Rating Scale (range 0–60, with higher scores indicating more depressive symptoms), and a variable indicating whether caregivers received informal/formal care for their own needs. Caregiving-specific variables included: hours of care provided in the last month, categorized into three groups (<1h, 10% of caregivers; 1–30h, 60% of caregivers; ≥30h, 30% of caregivers), interruptions of sleep during the night (number of nights per week), frequency of perceived burden and limitations in personal life due to caregiving (never, rarely-sometimes, often-almost always), and type of caregiver-care receiver relationship (spousal, non-spousal).

Two independent exposures were used in the multinomial models. The social isolation index was operationalized as a standardized Z-score derived from multiple indicators of social connections, including marital status, number of children, and frequency of social contacts. Each component was standardized and then combined into a single index, with higher scores reflecting higher levels of social isolation, as previously described [23]. For this study, social isolation was categorized into tertiles (low, moderate, high), and also used as a continuous variable in interaction analyses. Loneliness was determined by asking a single-item question “How often do you feel lonely?” and was categorized into three reported frequency levels (never, rarely, sometimes/often). These exposures were considered and modelled independently, since they measure different social constructs and are often weakly correlated (Cramér’s V = 0.16 in our dataset).

A total of 994 unique caregivers were providing care at some point during the study period, leading to a total of 1382 observations. The selection of the study sample is summarized in the study population flowchart (Fig. 1). The mean age of caregivers in SNAC-K was 73 years (SD 8.3) and 66% were female.

The outputs of the MCA and cluster analysis are shown in Fig. 2. The first two dimensions of the MCA explained 85.9% of the data variability, 79.8% and 6.1%, respectively. The contribution of each variable and response options to both dimensions is presented in Fig. 2A. Three clusters—good QoL (57.9% of the sample), moderate QoL (34.8% of the sample) and moderate physical, poor mental QoL (7.3% of the sample)—were identified by the k-means method, with a good cluster stability as indicated by the Jaccard bootstrap coefficients: 0.96, 0.97, and 0.93, respectively. The representation of the clusters, together with the QoL items, are shown in Fig. 2B. Differences in each item of the QoL scale are presented in Supplementary Table 2.

A description of the total sample and the three QoL profiles is presented in Table 1. The moderate physical, poor mental QoL profile was characterized by an overrepresentation of women, university-educated and unmarried persons, with inadequate physical activity and high depression scores, who provided the most intense care (in hours), often to a spouse. They reported the highest limitations to life and feelings of burden due to caregiving.

Compared with caregivers in the good QoL profile, those in the moderate physical, poor mental QoL profile were 3.2 times (95% CI: 1.6–6.3) more likely to belong to the moderate social isolation tertile and 5.6 times (95% CI: 2.7–11.6) more likely to belong to the high social isolation tertile. They also had a 1.7-times (95% CI: 0.9–3.5) and 4.8-times (95% CI: 2.6–9.0) higher risk of reporting loneliness as rarely or sometimes/often (as opposed to never), respectively, compared to those in the good QoL profile (Table 2). No significant associations were seen for the moderate QoL vs. good QoL profile comparison. We observed a significant interaction between sex and social isolation (p = 0.02). As shown in Table 2; Fig. 3, the odds of belonging to the moderate physical, poor mental QoL profile increased more steeply with higher levels of social isolation among males than among females, indicating a stronger association between social isolation and poorer QoL in men.

This study offers new insights into the QoL of older informal caregivers in Sweden, highlighting key factors that affect their well-being. We identified three distinct QoL profiles: good, moderate, and moderate physical, poor mental QoL. While most caregivers fell into the good or moderate QoL profiles, some were categorized in the moderate physical, poor mental QoL profile. While it is well established that loneliness and social isolation adversely affect older adults’ well-being, our study extends this knowledge by showing that, within the caregiving population, these social factors are specifically linked to the mental rather than physical dimensions of QoL. By adopting a data-driven clustering approach, we identified distinct caregiver profiles instead of treating QoL as a single continuum, providing a more nuanced understanding of how well-being dimensions co-occur.

The descriptive analysis of the three QoL profiles in this study highlights the significant heterogeneity among older caregivers, whose own health strongly influences their well-being. Caregivers in the poorest QoL profile (moderate physical, poor mental), who were more likely to be women, spousal caregivers, and those providing more hours of care, had more depressive symptoms and were frequently in need of care themselves. These findings are not surprising and align with previous literature, which shows that older caregivers often have their own health needs that must be addressed in addition to their caregiving responsibilities [9]. The overrepresentation of women and spousal caregivers in the worst QoL profile further emphasizes the challenges these caregivers face, particularly given their higher caregiving intensity and increased emotional and physical strain [25, 26]. Our findings add to this literature by empirically demonstrating the coexistence of poor physical and mental QoL within a subset of caregivers, reinforcing the need to view caregiver well-being as a multidimensional concept.

Loneliness and social isolation emerged as significant factors associated with the worst QoL profile. Previous research suggests that spousal caregivers are at risk of social isolation because they often feel ashamed of their care recipient’s behaviors or are simply too busy to maintain social connections [27, 28]. Additionally, the caregiving burden can lead to a reduction in leisure time, further exacerbating isolation [29]. We also found that the most significant differences in QoL were between the worst and best profiles, rather than between the middle and best profiles. This suggests that it is primarily the mental component of QoL that is correlated with loneliness and isolation. Caregivers in the moderate profile still reported relatively high mental QoL, indicating that they are better able to manage the psychological challenges associated with caregiving. In contrast, caregivers in the worst profile, with poorer mental health, were more vulnerable to experiencing social isolation and, subsequently, the many negative health outcomes it has been linked to.

The observed interaction between sex and social isolation provides an additional novel insight. While women were overrepresented in the poorest QoL group, suggesting higher caregiving intensity, men with similarly high levels of social isolation were more likely than women to belong to this worse profile. This suggests that men may be particularly vulnerable to the psychological consequences of isolation, potentially due to smaller social networks, a greater reliance on spouses for social connection, or less access to emotional support [30, 31]. Conversely, women may be somewhat better able to cope with isolation, possibly through stronger relational or coping resources. These findings highlight that sex not only shapes caregiving roles but also moderates how social isolation translates into mental well-being.

This study has several key strengths. The use of robust statistical methods, including multiple correspondence and cluster analyses, allowed for a detailed identification of caregiver profiles based on QoL measures. Additionally, the study benefits from rich, reliable data from an ongoing cohort study. A major strength is the focus on older informal caregivers, an often-overlooked group, offering valuable insights into their challenges.

However, there are several limitations. The cross-sectional design limits the ability to establish temporality, making it unclear whether worse QoL leads to or results from loneliness and social isolation. Most likely, this association is bidirectional. The lack of information about care receivers (e.g., their health status or care needs) limits understanding of the caregiver/care receiver dynamics. Additionally, our dataset included 11 of the 12 items from the SF-12 tool, with item 7 not available in the original data collection. Although this omission slightly reduces the completeness of the standard SF-12 instrument, its impact on our findings is likely minimal. First, summation of raw item scores is a recognized method for operationalizing SF-12 constructs [19]. Second, because each SF-12 item was analyzed separately in the MCA, rather than being aggregated into summary scores, the absence of a single item does not distort the clustering structure or bias the identification of QoL profiles. Finally, the cohort for this study was drawn from Kungsholmen, a central island in Stockholm, Sweden, characterized by a predominantly white and middle-to-high-income population. This limits the generalizability of the results to more ethnically diverse and/or low-income settings.

This study highlights the heterogeneity in QoL of older informal caregivers in Sweden, revealing three distinct profiles: good, moderate, and moderate physical, poor mental QoL. While most caregivers reported good or moderate QoL, a notable subset experienced poor mental health, and belonging to the latter group was strongly associated with loneliness and social isolation. Women, spousal caregivers, and those providing intensive care were also overrepresented in this group, underscoring the substantial emotional and physical strain they face. The findings emphasize the critical role of social well-being in maintaining caregivers’ QoL, with sex-specific differences highlighting men’s heightened vulnerability to isolation. Tailored support services, equitable access to resources, and regular evaluations of caregiver interventions are necessary to address these challenges. Future research should focus on the longitudinal and intersectional factors influencing caregivers’ QoL to guide evidence-based policies that support their well-being and recognize their vital societal contributions.