Are socially isolated people more likely to develop cancer? Study of 350,000 adults explores the risk

By Vijay Kumar MalesuReviewed by Susha Cheriyedath, M.Sc.Mar 3 2026

A massive UK Biobank analysis suggests that objective social isolation may modestly increase cancer risk, particularly in women, highlighting how social conditions and lifestyle factors could shape long-term cancer outcomes.

Study: A study of the associations between social isolation and loneliness with sex-specific cancer risk in the UK Biobank. Image Credit: Halfpoint / Shutterstock

In a recent study published in the journal Communications Medicine, a group of researchers examined whether social isolation and loneliness are independently associated with overall and site-specific cancer incidence, while also evaluating sex differences and potential biological and behavioral pathways that may contribute to these associations.

Background

Nearly one in four people report feeling socially isolated at some point in life, and loneliness has been described as a growing public health concern worldwide. Beyond mental health, researchers now question whether limited social connections may influence chronic diseases such as cancer.

Social isolation differs from loneliness, which reflects a subjective feeling of being alone. Social isolation and loneliness are often associated with inflammation, unhealthy behaviors, and earlier death. Yet whether they increase the risk of developing cancer remains uncertain. Understanding this link matters because social relationships are potentially modifiable. Further large-scale prospective research is needed to clarify these associations.

About the study

This prospective cohort study used data from the UK Biobank, which recruited over 500,000 adults aged 38 to 73 years between 2006 and 2010. After excluding participants with missing exposure data or those diagnosed with cancer within one year after baseline, 354,537 individuals remained in the analytic cohort. Social isolation was measured using three factors, including living alone, infrequent social visits, and lack of weekly social participation. Participants who scored 2 or more points were classified as socially isolated. Loneliness was defined using two questions: one assessing frequent feelings of loneliness and the other asking whether participants reported being unable to confide in someone close.

Cancer incidence was identified from national registries using ICD-10 codes C01-C97, excluding non-melanoma skin cancer. Demographic, economic, physical activity, health status, and psychological characteristics were adjusted for while applying Cox proportional hazards models and Fine-Gray competing risk models. Complete blood cell counts, C-reactive protein, neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, lymphocyte-to-monocyte ratio, systemic immune-inflammation index, and other inflammatory biomarkers were also evaluated as potential mediators of inflammation-related pathways. Subgroup and sensitivity analyses were conducted to test robustness.

Study results

Over a median follow-up of 11.6 years, 38,103 participants developed cancer. At baseline, 5.8% were socially isolated, and 4.5% reported loneliness. Socially isolated persons were more likely to have less income, lower education, higher body mass index, poorer sleep patterns, and higher smoking rates, all of which are risk factors for cancer. After full adjustment, social isolation was associated with an approximately 8–9% higher risk of developing cancer (cause-specific hazard ratio approximately 1.09), whereas loneliness showed no independent association. Cancer incidence increased across categories reflecting greater combined exposure to social isolation and loneliness; however, loneliness alone was not associated with overall cancer risk after adjustment, and no statistical interaction between social isolation and loneliness was observed. Importantly, sex differences emerged in stratified analyses, and social isolation showed a significant association with cancer risk among females but not consistently among males after adjustment.

Among women, social isolation was associated with a higher incidence of breast, lung, uterine, ovarian, bladder, and stomach cancers. Among men, the most significant finding was an association between social isolation and bladder cancer incidence. These findings are notable, as breast and lung cancers are two of the most common cancers in the world today. The results suggest that women experiencing limited social connections may represent a potentially underrecognized group with elevated cancer risk, although these site-specific findings were derived from multiple comparisons and should be interpreted cautiously.

Mediation analyses indicated that a substantial proportion of the association between social isolation and cancer risk was statistically explained by socioeconomic disadvantage, unhealthy behaviors (smoking, alcohol consumption, poor diet, low physical activity, poor sleep), and poorer overall health. Inflammatory markers were estimated to account for a smaller portion of this association, particularly neutrophil count and white blood cell count. For example, neutrophils explained approximately 9% of the excess risk in the overall population. In women, inflammatory pathways also statistically mediated part of the associations for breast and lung cancers, and exploratory analyses in the study also evaluated hormonal factors as potential contributors to sex differences.

Interestingly, loneliness alone did not increase overall cancer risk. In some subgroups, such as younger or employed individuals, loneliness was associated with a slightly lower observed cancer risk, suggesting complex social and psychological dynamics. These patterns suggest that objective social disconnection, rather than subjective feelings alone, may exert stronger behavioral or physiological effects, though these subgroup findings require cautious interpretation.

Sensitivity analyses excluding early cancer cases and accounting for competing risks produced similar results, supporting the findings.

Conclusions

Social isolation, but not loneliness alone, was associated with a modest yet meaningful increase in cancer incidence, particularly among women. Socioeconomic disadvantage, health-related behaviors, and inflammatory markers accounted for part of this association in statistical mediation analyses. These findings underscore that cancer risk may be influenced not only by genetic and medical factors but also by social conditions and behavioral pathways. However, because this was an observational study, the findings demonstrate associations rather than causation.

Additionally, because the UK Biobank cohort consists predominantly of middle-aged and older adults of European ancestry and may reflect a “healthy volunteer” population, the generalizability of these findings to more diverse populations may be limited. Future interventional and mechanistic studies will be needed to determine whether reducing social isolation can meaningfully influence cancer risk or long-term health outcomes.