What does loneliness actually do to health?
It is associated with measurably worse physical health outcomes, not only unhappiness — and the size of the association has surprised researchers enough that several public health bodies now treat it as a health issue rather than a social one.
What the evidence shows. Meta-analyses have found that social isolation and loneliness are associated with substantially increased risk of early death, with effect sizes frequently compared to well-known risk factors such as smoking and obesity. The comparison is widely quoted and is worth stating carefully: these are observational associations, and disentangling cause from effect is genuinely difficult, since poor health also causes isolation.
The proposed mechanisms:
Chronic stress response. Loneliness is associated with elevated stress hormones and with markers of chronic low-grade inflammation, which is implicated in cardiovascular disease, diabetes and dementia.
Immune changes, including altered responses to vaccination and infection in some studies.
Sleep quality, which is measurably worse in lonely people even when total sleep time is similar.
Behaviour. Isolated people exercise less, eat less well, drink more and are less likely to seek medical help or adhere to treatment — which accounts for a meaningful part of the association.
Fewer people noticing, which matters practically for early detection of problems.
The important distinction. Loneliness is the subjective feeling; isolation is the objective circumstance. They frequently coexist and do not have to. Someone can be alone and content, or surrounded by people and lonely — and the subjective experience predicts health outcomes at least as strongly as the objective one.
What helps, according to intervention research. Simply adding social contact works less well than expected. The approaches with better evidence address the thinking that maintains loneliness — lonely people become hypervigilant to rejection and withdraw pre-emptively — alongside opportunities for contact built around shared activity rather than socialising as the purpose, which is why groups organised around doing something outperform groups organised around meeting.
Who is most affected: not only older people — young adults report high rates, and transitions such as moving, bereavement, caring and retirement are the reliable triggers.