Loneliness Raises the Risk of Early Death by About a Quarter, and the World Health Organization Now Counts It as a Cause

In June 2025 the WHO Commission on Social Connection published its first global report, estimating that one person in six worldwide is affected by loneliness and that it is linked to about 871,000 deaths a year, roughly a hundred an hour. The rates were highest among adolescents and young adults and in low-income countries, not, as the stereotype has it, among the elderly in rich ones.

How it does harm

The mechanisms are partly behavioural and partly physiological. Lonely people sleep worse, drink more, exercise less and are less likely to see a doctor or take medication as prescribed. Independently of that, chronic loneliness is associated with raised cortisol, higher blood pressure, elevated inflammatory markers and altered immune gene expression, a pattern consistent with a body under sustained low-grade threat. Loneliness is also a strong predictor of depression and of cognitive decline in later life. None of these pathways is fully proven to be causal, and some of the association certainly runs the other way, since illness isolates. But the consistency across studies, countries and measures, and the persistence of the effect after adjustment, has convinced most epidemiologists that a substantial part of it is real.

The caveats

The comparison with smoking is a comparison of effect sizes in pooled observational studies, not a statement that a lonely person and a smoker face the same risk. The studies measure loneliness in dozens of different ways. The living-alone figure is confounded by everything that determines who lives alone. And an increase in relative risk of a quarter, applied to a low baseline risk in a healthy forty-year-old, is a small absolute number, though applied across a population it is a great many deaths. What the evidence supports is that persistent loneliness is a genuine health risk of roughly the same order as several the medical system already takes seriously, and that it is common.

What has been shown to help

Less than one would hope, and the honest summary is that the field is young. Reviews of interventions consistently find that programmes which give people a shared activity and a role, volunteering, group exercise, choirs, classes, adult education, outperform programmes that simply arrange contact, such as befriending visits. Structure and purpose seem to matter more than company alone. Cognitive approaches that address the way lonely people come to expect rejection and withdraw pre-emptively have some of the better trial evidence. Technology has a mixed record: video calls with family help; passive scrolling does not, and may hurt.

For the individual, the practical implications are modest and old-fashioned. Regular, scheduled contact beats occasional contact. Doing something alongside someone beats talking to them. Small, repeated interactions with neighbours, shopkeepers and colleagues, the weak ties that sociologists have long known matter, carry real weight. And loneliness, like pain, is a signal, not a character flaw; treating it as information rather than as shame is the first step most programmes try to teach.

Why this matters more here

The Gulf has an unusually high share of people living far from their families, often for years, often alone or in shared accommodation with strangers, and often working hours that leave little room for anything else. Our coverage of watching family in a conflict zone from abroad and of managing stress through a regional crisis touched on this population. The loneliness evidence is a reason to take their situation seriously as a matter of health rather than only of welfare, and it is a reason for anyone in that position to treat a weekly call, a Friday football game or a shared meal as maintenance rather than as luxury.