A nine-year study reveals a pattern linking loneliness to heart risk

By Dr. Liji Thomas, MDReviewed by Lauren HardakerOct 21 2026

Social isolation may not tell the whole story. A nine-year study of more than 10,000 adults reveals how changing feelings of loneliness, even among those with moderate social isolation, are linked to a higher risk of heart disease and stroke.

Study: Co-occurring patterns of social isolation and loneliness and their associations with cardiovascular disease risk among middle-aged and older adults: a group-based dual-trajectory modeling study. Image credit: WPixz/Shutterstock.com

A recent study published in Archives of Public Health found that certain patterns of social isolation and loneliness over time were associated with cardiovascular disease (CVD) risk among middle-aged and older adults.

Why loneliness and social isolation matter for heart health

Social isolation and loneliness are related but distinct psychosocial factors. Social isolation refers to having few social relationships or limited contact with others. Conversely, loneliness is the subjective feeling that one's social relationships are insufficient or do not meet one's needs.

Both factors have been associated with adverse health outcomes, including cardiovascular disease. Individual or cross-sectional examination of these factors may not capture their interactions and changes over time.

In this study, the researchers aimed to identify longitudinal patterns of social isolation and loneliness occurring together, and investigate their associations with subsequent CVD risk among adults aged 45 years and older.

Tracking loneliness and social isolation over nine years

The study included 10,823 participants from the China Health and Retirement Longitudinal Study (CHARLS), a longitudinal study of middle-aged and older adults in China. The researchers analyzed data collected between 2011 and 2020. Participants with CVD at baseline were excluded from the analysis.

Social isolation was assessed using five indicators, including marital status, living arrangements, contact with children, social participation, and rural residence. Loneliness was measured using a single question about how frequently participants had felt lonely during the previous week.

They used group-based dual-trajectory modeling to cluster participants based on similar patterns of co-occurring social isolation and loneliness over time. Unlike an analysis based on a single measurement, this approach allowed the researchers to examine how the two factors developed together.

The researchers then used discrete-time proportional hazards models to examine the associations between these patterns and CVD risk. Participants in the group characterized by moderate social isolation and low loneliness served as the reference group for comparisons.

Social isolation and loneliness were assessed repeatedly during follow-up, while cardiovascular events were recorded over the same study period. Consequently, the analysis could not always establish whether changes in social relationships preceded the onset of cardiovascular disease.

The analysis was observational. Therefore, the study could identify associations between the trajectories and CVD risk but could not establish that social isolation or loneliness caused cardiovascular disease.

Patterns of social isolation and loneliness

The analysis identified five distinct patterns of social isolation and loneliness. During follow-up, the analysis reported 2,589 new CVD cases. The highest CVD risk compared to the reference group was in the group with moderate social isolation and fluctuating loneliness, with a 53% increase in risk. This association suggests that even without high social isolation, changes in loneliness over time may be associated with higher CVD risk. In this group, loneliness initially declined, increased markedly between 2013 and 2018, and then declined again by 2020.

The risk was also elevated among participants with both high social isolation and high loneliness, at 29% increased CVD risk relative to the reference group.

However, the risk was also 16% higher among participants with low social isolation and low loneliness. Social isolation increased in this group during the later follow-up years, and participants had higher rates of smoking, alcohol consumption, and overweight or obesity. The researchers cautioned that unmeasured health factors and other sources of bias could also help explain this unexpected association. This result indicates complex associations between these risk factors and CVD risk, rather than a linear pattern.

Participants with high social isolation but low loneliness did not have a statistically significant difference in CVD risk compared with the reference group, which had moderate social isolation and low loneliness.

Notably, the lack of a statistically significant difference in CVD risk among participants with high social isolation but low loneliness stood in contrast to the elevated risk associated with fluctuating or high loneliness. However, this finding does not establish that high social isolation is harmless or that low loneliness protects against cardiovascular disease.

This supports the importance of distinguishing a person's objective social circumstances from their subjective experience of those circumstances.

Why the findings cannot establish cause and effect

These findings should be interpreted cautiously. This was an observational study and cannot establish causality or determine the preventive value of interventions to reduce loneliness. Other factors may contribute to both loneliness and CVD risk, confounding the associations. Poor underlying health may also have contributed to social isolation or loneliness before cardiovascular disease was diagnosed, raising the possibility of reverse causation.

The study assessed loneliness using a single question, which may not capture its full complexity. Cardiovascular outcomes relied on self-reported physician diagnoses of heart disease or stroke rather than independent clinical verification, introducing the possibility of measurement errors.

The study sample was Chinese, limiting generalizability. Additionally, the social isolation index included rural residence, a context-specific measure that may not be applicable in other populations or directly comparable with other studies.

Changing loneliness patterns may help identify cardiovascular risk

The study suggests that the relationship between social isolation, loneliness, and cardiovascular disease is more complex than a simple association between the size of a person's social network and their health. Considering both factors together may be more useful than either alone.

The strongest association was observed among participants with moderate social isolation and fluctuating loneliness. Another vulnerable group experienced both high social isolation and high loneliness. The researchers suggest that such combined assessments could help target at-risk middle-aged and older adults for cardiovascular risk screening combined with social or emotional support where appropriate. However, further research is needed to establish whether improving social relationships can reduce cardiovascular disease risk.