Isolation and Loneliness Together Track Higher Cardiovascular Risk in Older Adults

People whose loneliness rises and falls over the years, while their social isolation stays moderate, face the highest risk of cardiovascular disease among middle-aged and older adults, according to a new modeling study drawing on nearly a decade of data from China. The finding suggests that neither isolation nor loneliness alone tells the full story, and that the way the two evolve together over time may matter for heart health.

Social isolation and loneliness are related but distinct conditions.

In the new study, published in Archives of Public Health, a team led by Zhengkun Liu and Chunxia Cao analyzed data from 10,823 adults aged 45 years and older who took part in the China Health and Retirement Longitudinal Study between 2011 and 2020. The dataset allowed the researchers to follow participants across nine years, tracking repeated measurements of both social isolation and loneliness for each individual.

The method at the heart of the paper is group-based dual trajectory modeling, a statistical technique that identifies hidden subgroups of people who share similar longitudinal patterns on two measures simultaneously. Instead of asking whether isolation or loneliness predicts disease at one snapshot, the approach classifies participants according to how both measures unfold together over the follow-up period, producing joint trajectory groups that reflect the combined course of the two psychosocial states.

That analysis revealed five distinct co-occurring patterns among the middle-aged and older adults in the sample. Each pattern combined a characteristic level of social isolation with a characteristic course of loneliness, ranging from low levels of both to high levels of both, with intermediate groups showing moderate isolation paired with either low or fluctuating loneliness.

To link these patterns to disease, the researchers used discrete-time proportional hazards models, a survival-analysis framework suited to data collected in discrete intervals such as survey waves. This allowed them to estimate how the risk of cardiovascular disease differed across the trajectory groups while accounting for the passage of time during follow-up.

The results pointed clearly to one group as the most vulnerable. Compared with participants in Pattern 2, who had moderate social isolation and low loneliness, those in Pattern 4, marked by moderate social isolation and fluctuating loneliness, had the highest cardiovascular disease risk, with a hazard ratio of 1.53, a 95 percent confidence interval of 1.36 to 1.72. Elevated risks also appeared in Pattern 5, characterized by high social isolation and high loneliness, with a hazard ratio of 1.29 and a 95 percent confidence interval of 1.10 to 1.49, and in Pattern 1, with low social isolation and low loneliness, which showed a hazard ratio of 1.16 and a 95 percent confidence interval of 1.04 to 1.30.

Not every departure from the reference group carried risk. Pattern 3, in which participants had high social isolation but low loneliness, did not differ significantly from Pattern 2 in cardiovascular disease risk. That asymmetry is one of the study’s most striking implications: objective isolation alone, when not accompanied by rising or high subjective loneliness, was not associated with elevated risk in this analysis, whereas loneliness that fluctuated or remained high appeared to matter more, particularly when combined with moderate or high isolation.

The authors conclude that identifying middle-aged and older adults at elevated cardiovascular risk may require looking at the joint longitudinal patterns of isolation and loneliness rather than either factor in isolation. They suggest that individuals with fluctuating or high loneliness, especially alongside moderate or high social isolation, may warrant closer cardiovascular risk assessment and connection to social or emotional support resources. As an observational analysis of a single national cohort, the study identifies associations between psychosocial trajectories and disease risk over nine years of follow-up; the five-group structure and the hazard estimates describe this specific population and modeling approach, and the findings highlight a measurable pattern worth attention rather than establishing that changing loneliness would directly alter cardiovascular outcomes.

Subject of Research: Co-occurring longitudinal patterns of social isolation and loneliness and their association with cardiovascular disease risk in middle-aged and older adults

Article Title: 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

Article References: Liu, Z., Sun, X., Li, Y., Mei, Z., Yuan, J., & Cao, C. (2026). 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. Archives of Public Health. https://doi.org/10.1186/s13690-026-02102-y

Image Credits: AI Generated

DOI: 10.1186/s13690-026-02102-y

Keywords: social isolation, loneliness, cardiovascular disease, CHARLS, trajectory modeling, older adults, middle-aged adults, hazard ratio, longitudinal study, psychosocial factors, China, public health