Men who were both lonely and socially isolated in midlife could expect about 5.8 fewer years free of depression and anxiety from age 50 than men who were neither, and women about 3.7 fewer years, according to an analysis of 277,489 adults in the UK Biobank led by researchers at Tulane University. The peer-reviewed study was published in Nature Communications on August 20, 2026; Tulane publicized it on September 28 and ScienceDaily carried the release this month.

Participants were aged 40 to 69 when recruited between 2006 and 2010 and were free at baseline of seven conditions: type 2 diabetes, cardiovascular disease, chronic respiratory disease, neurodegenerative disease, cancer, depression and anxiety. They were followed through December 31, 2022, a median of 13.4 years for new diagnoses and 13.8 years for deaths, during which 78,829 developed one of the seven conditions and 16,356 died, the paper reports.

Loneliness was defined by two baseline questions: whether participants often felt lonely and whether they could confide in someone close less than once a month; a person was classed as lonely if both applied. Social isolation was scored on three items, living alone, visiting friends or family less than monthly, and taking part in no weekly group activity, with two or more counting as isolated. As Tulane’s announcement put it, isolation is an objective measure of how much social contact a person has, while loneliness is a subjective dissatisfaction with the quality of one’s relationships.

To translate risk into years, the researchers built a multistate life table with three states, disease-free, diseased and dead, using sex- and age-specific transition rates and hazard ratios from Cox models, and ran the tables from age 50 to 100 with 2,000 bootstrap replications to produce confidence intervals. Physical disease-free life expectancy covered the five physical conditions; mental disorder-free life expectancy covered depression and anxiety.

Compared with people who were neither lonely nor isolated, women with both lost 2.4 years free of physical disease and 3.7 years free of mental disorder at age 50; men lost 1.7 and 5.8 years. Each exposure alone also mattered. Loneliness on its own was associated with 3.9 fewer mental disorder-free years in women and 2.9 in men; isolation on its own with 2.0 and 3.0 fewer years, respectively. Losses in physical disease-free years ranged from 0.9 to 1.5 years for either exposure alone.

A healthy lifestyle, scored on smoking, body mass index, physical activity, diet and sleep, narrowed the loss associated with social isolation, but the corresponding association for loneliness was not statistically significant.

Lead author Lu Qi, HCA Regents Distinguished Chair and professor at Tulane’s Celia Scott Weatherhead School of Public Health and Tropical Medicine, said in the announcement: “The findings weren’t surprising to me.” On the sex difference, he said: “Our research found men are more susceptible to being affected by loneliness.” The work was funded by the US National Institutes of Health and Fudan University, and the authors declared no competing interests.

Why it matters

Loneliness and isolation are widespread: Tulane cites a 2023 report that roughly half of US adults are lonely and nearly as many are socially isolated. The study’s contribution is to express the associated harm in a unit people and policymakers understand, years of life without a diagnosis, and to show that the toll falls more heavily on mental health than on physical health, and more heavily on men than on women where mental disorders are concerned.

The lifestyle finding has practical implications. Diet, exercise, sleep and not smoking appear to blunt the harm linked to isolation but not the harm linked to loneliness, which the authors say suggests the two conditions may need different interventions. The researchers plan to examine other population groups and to test causality in clinical trials, according to the announcement and a report in News-Medical.

The caveats

This is an observational study and cannot establish that loneliness or isolation cause depression, anxiety or physical disease. Both exposures were measured once, at baseline, with brief questionnaires, which the authors say may underestimate the true associations. The UK Biobank enrolled about 5.5 percent of those invited and is not representative of the general population, lifestyle was self-reported, and the model’s assumption that transitions depend only on the current state was not formally tested. The headline figure of nearly six years applies to men with both exposures, a subgroup, not to lonely people in general.

This article is for general information and does not replace advice from a qualified clinician. Corrections: email the editor using the address on the Contact page.