Social isolation is defined as “the objective lack of (or limited) social contact with others,” typically characterized by few social ties, infrequent interactions, or living alone [1, 2]. It is measured using objective indicators—such as number of contacts, frequency of interaction, and network size—that capture the structural aspects of social networks. Although self-reported, these indicators are considered objective as they capture quantifiable characteristics of social connections rather than subjective perceptions. Social isolation is a structural network characteristic of social networks rather than loneliness or perceived support. It is used to assess how the absence of social contacts contributes to mortality risk independently of subjective experiences.

Social isolation differs from loneliness, which is a subjective feeling of being isolated [2]. Social support refers to the resources provided by others in the social network [3]. Although isolation and loneliness often overlap, they are not interchangeable: some individuals may be isolated without feeling lonely, while others may feel lonely despite having many social connections [4].

Social relationships play a crucial role in both physical and mental health by providing emotional and instrumental support (e.g., practical help with daily tasks), shaping health behaviors, buffering stress, and facilitating access to care [5]. Through these mechanisms, social relationships enhance psychological wellbeing, reduce the risk of depression, and contribute to physiological regulation, including immune and neuroendocrine function [5].

Conversely, the absence of social relationships has emerged as a major public health concern. Social isolation has been found to have health effects comparable to alcohol consumption and tobacco use [6–8], and its prevalence increases with age [9]. The mechanisms linking social isolation to mortality can be understood through behavioral, psychological, and physiological pathways. Isolation reduces access to social support and engagement, increasing the likelihood of unhealthy behaviors such as physical inactivity, poor diet, and low adherence to medical care. Physiologically, isolation has been associated with chronic inflammation, immune suppression, and dysregulation of the hypothalamic–pituitary–adrenal axis. Psychosocially, isolation contributes to chronic stress, depression, and diminished self-efficacy and self-esteem. These pathways are illustrated in the social epidemiology framework which situates social networks as upstream social-structural determinants that shape psychosocial processes and, in turn, health outcomes [10]. Within this framework, our study examines social isolation as a structural network characteristic and its association to all-cause and cause-specific mortality among older Swedish adults.

Non-communicable diseases, particularly cancer and cardiovascular disease, are the leading causes of death among older adults [11, 12]. In Europe, Cardiovascular Disease (CVD) remains the top cause of mortality in both men and women, followed by cancer [13–15]. In Sweden, ischemic heart disease (IHD) is the most common cause of death in men and the second most common in women, while Alzheimer’s disease and other dementias are the leading cause of death in women [12, 16]. Regarding cancer, prostate cancer is the leading cause of cancer-related death among men, whereas lung cancer is the leading cause among women, followed by breast cancer [12, 17]. Given the substantial burden of these diseases, identifying factors such as social isolation that may contribute to mortality is critical for developing preventive strategies and reducing premature mortality among older adults.

A growing body of research consistently demonstrates an association between social isolation and a higher risk of mortality in diverse populations. For example, this association was shown in studies of African-Americans aged 20–90 years [18], individuals with obesity in the UK Biobank cohort [19], older adults in Japan and in England [20], and adults over 50 in Mexico [21]. Notably, one recent study found that social isolation increased 20-year mortality risk by more than 15% [22]. Furthermore, findings from the Prospective Urban Rural Epidemiology (PURE) study revealed that social isolation was associated with an increased risk of all-cause mortality among adults aged 35–70 across countries with differing socioeconomic conditions [23].

Research examining cause-specific mortality has shown that socially isolated individuals face a higher risk of death from CVD [24] and cancer [25, 26]. A meta-analysis further demonstrated that social isolation significantly increases the risk of both CVD and cancer, specifically breast cancer [18]. Among individuals already diagnosed with cancer, social isolation has also been linked to increased mortality [27].

Sweden is a particularly relevant context for studying social isolation, as it has one of the highest proportions of older adults living alone globally, a trend that has increased over time [28, 29]. Older adults living alone are at a higher risk of experiencing loneliness and social isolation, and a recent study found that approximately 31% of men and 59% of women aged 75 and older in Sweden live alone [29]. Among this group, 52.6% of men and 73.7% of women reported feeling lonely [30].

A previous Swedish study with a 5-year follow-up found that both social isolation and loneliness were associated with increased mortality risk [26]. However, the study had several limitations: first, social isolation was measured using limited indicators that did not capture the full range of social contacts and activities; second, relatively few participants experienced both high isolation and high loneliness, limiting analysis of their combined effects; and third, reverse causality was a concern, as poor health could lead to social isolation. Unlike prior research, our study addresses these limitations by employing a longer follow-up period of up to 10 years, thereby strengthening causal inference. We also use a broader set of indicators to capture social isolation as a structural network characteristic, providing a more comprehensive assessment of social connections. Moreover, our large population-based cohort increases the statistical power to explore associations with both all-cause and cause-specific mortality.

This study examines the association between social isolation and both all-cause and cause-specific mortality among Swedish older adults aged 56–95 years, by linking comprehensive questionnaire survey data to the Swedish Cause of Death register. By including cause-specific mortality, this study seeks to provide more detailed insights into how social isolation affects health outcomes, extending beyond the scope of previous studies.

This study received ethical approval from the Swedish Ethical Review Authority (DNR- 2019-03643).

The participants had an average age of 68.75 years (SD = 7.85), ranging from 56 to 95 years, and just over half (55.91%) were men, while just under half were women (44.09%). The mean follow-up time was 10.17 years (SD = 2.05). Chronic stress in private life was reported by a small proportion (5.89%) of the sample. Nearly all participants (99.24%) lived in their own homes, with very few residing in assisted living or retirement facilities. Approximately one-quarter of the participants lived alone. Table 1 presents baseline characteristics of the study population.

Nearly half of the participants did not engage in physical activity, while 55.47% reported some form of exercise. A large majority (88.02%) were current drinkers, while about half (50.76%) had never smoked, more than a third (39.98%) were former smokers, and nearly one in ten (9.26%) were current smokers. Social relationships, a key focus of the study, were assessed using the LSNS-6 scale to measure social isolation. Based on this measure, about the same proportion of participants were categorized as being at high risk (40.83%) and moderate risk (39.11%) of social isolation, while about one in five (20.06%) at were categorized as low risk. Club or social activity participation was relatively low, with about half (49.82%) of participants not engaged in such activities, highlighting the potential difficulties they may face in maintaining strong social networks.

During the follow-up period, about one in six (17.43%) participants had died, while the majority (82.57%) of participants were censored. The overall mortality rate was 17.14 deaths per 1,000 person-years. Table 2 presents the primary causes of death. CVD and cancer were the most common causes of death. Within CVD, acute myocardial infarction (AMI) was a leading cause, with 467 deaths and a mortality rate of 173.24 per 1,000 person-years. Ischemic heart disease (IHD) followed with 492 deaths and a mortality rate of 155.61 per 1,000 person-years. Among cancer-related deaths, breast cancer was a significant cause of death among women, with 102 deaths (1.60%) and a mortality rate of 177.04 per 1,000 person-years. Prostate cancer was a major cause among men, with 336 deaths (5.28%) and a mortality rate of 162.58 per 1,000 person-years.

Our study found a significant association between social isolation and both all-cause and cause-specific mortality among older adults in Sweden. Among men, both moderate and high social isolation were associated with increased mortality hazards, whereas in women, the association was only seen for high social isolation.

These findings align with previous research showing an association between social isolation and increased all-cause mortality in older adults aged 69 years and over in Sweden [26]. Living alone, often referred to as a proxy indicator of social isolation [37], is common in Sweden and has been significantly linked to higher mortality among older adults in Västerbotten County [38]. Studies from other settings showed a similar association [19–21, 23], while higher levels of loneliness have been linked to reduced social support and increased social isolation [39]. Meta-analyses further confirm that social isolation is associated with an elevated risk of mortality [7, 40]. Our findings confirm that social isolation is a predictor of mortality, comparable to other well-established risk factors such as smoking [8].

In the current study, we did not examine potential pathways from social isolation to mortality. However, several mechanisms have been proposed that may explain these relationships. Social isolation has been linked to chronic stress and inflammation [41–43], accelerated aging [44], unhealthy behaviors such as physical inactivity and poor diet [45–47], as well as poorer physical and mental health outcomes [48]. Together, these factors may contribute to the elevated risk of both all-cause and cause-specific mortality observed among socially isolated older adults.

In our analysis, high levels of social isolation were significantly linked to an increased risk of mortality from IHD. This observation is consistent with findings from a 17-year follow-up study of adults aged 18–64 [49] and longitudinal research showing that social isolation significantly heightens the risk of mortality among patients with IHD [50]. Stress-related physiological mechanisms likely play a role. Persistent activation of the hypothalamic–pituitary–adrenal axis and sympathetic nervous system elevates cortisol and catecholamines, which in turn promote endothelial dysfunction, vasoconstriction, platelet activation, reduced heart rate variability, and hypertension—factors that contribute to atherosclerosis formation and precipitate acute coronary events. [49, 51, 52]. Beyond to stress physiology, immune dysregulation may also play a role. Social isolation has been linked to higher levels of circulating Interleukin-6 (IL-6), CRP, and soluble urokinase plasminogen activator receptor (suPAR) [53, 54]. These inflammatory markers accelerate plaque formation and are strongly linked to cardiovascular mortality. In older adults, social isolation correlates with higher levels of IL-6 and CRP values, with elevated IL-6 in particular linked to increased cardiovascular mortality [55, 56].

We found that social isolation was associated with a higher risk of mortality from prostate cancer, corroborating findings from longitudinal cohort studies [57]. Previous research indicates that a lack of social support substantially influences prostate cancer stage at diagnosis and survival [58]. Psychosocial factors such as stigma, relational difficulties, and reduced social support may also delay help-seeking and reduce treatment adherence, ultimately leading to poorer outcomes [21, 59, 60]. In line with these findings, a Swedish cohort study reported that men with higher perceived stress not only had a significantly increased risk of prostate cancer–specific mortality (HR = 1.66; 95% CI: 1.05–2.63), but also reported fewer confidants and felt less able to share their problems with partners, family, and friends [61]. Beyond these psychosocial mechanisms, biological pathways may also be important: chronic stress-related activation of the hypothalamic–pituitary–adrenal axis and the autonomic nervous system may promote angiogenesis (new blood vessel formation) and tumor growth, thereby accelerating cancer progression and mortality [61, 62].

In the case of breast cancer, our study did not identify a significant relationship between social isolation and breast cancer mortality, consistent with research conducted in the United States [59]. This may be explained by high participation in the mammography screening program in Sweden [63], which effectively reduces mortality from breast cancer through early detection and treatment [64]. It is important to note that a Swedish study reported lower mammography screening participation among socially isolated women, highlighting potential disparities in access and utilization of preventive services [65]. Moreover, a large pooled cohort after the Breast Cancer Pooling Project revealed that socially isolated women faced a higher risk of breast cancer recurrence and breast cancer-specific mortality [66]. This discrepancy may reflect differences in study populations, the measurement of social isolation, or statistical power. While we did not investigate these specific sources of variation, they are important considerations when comparing results across studies.

Although social isolation and mortality have been studied in Sweden, evidence remains limited, particularly regarding cause-specific mortality. Our study addresses this gap by using a large sample size and an extended follow-up period, enabling a more thorough assessment of mortality risk over time while reducing potential bias.

This study also has limitations. First, exposure variables were self-reported, which may have introduced reporting bias. Second, social isolation was assessed only at baseline and did not account for changes over time. Future research should use repeated measures to capture the dynamic nature of social relationships. Third, 17,944 participants (33% of the initial sample) were excluded due to incomplete covariate data, raising the possibility of selection bias. Fourth, we did not include certain variables, such as club participation or pet ownership, which may plausibly influence social relationships and health but were not consistently measured. Fifth, we did not stratify by age or retirement status, which may influence patterns of social isolation and mortality. Finally, our analysis was limited to broad categories of cause-specific mortality, and the findings may not generalize to other causes of death or to populations outside Sweden.

In conclusion, social isolation was associated with all-cause mortality as well as cause-specific mortality from IHD in both sexes and prostate cancer in men, while no association was found for breast cancer mortality among older adults in Sweden. These findings highlight the urgent need for public health strategies that foster social engagement, particularly among older adults. Globally, policies addressing social isolation in aging populations should be prioritized, with an emphasis on fostering social integration through community-based programs. Further research is needed to clarify the complex mechanisms by which social isolation influences mortality and to inform the development of targeted interventions that can mitigate its adverse effects.