What does the evidence say about loneliness and lifespan?

How strong is the link between connection and lifespan?

Strong, and consistent across decades of work. The landmark meta-analysis pooled 148 prospective studies covering 308,849 people and found an odds ratio of 1.50 for survival among those with stronger social relationships. For richer measures of social integration the figure rose to 1.91.

Holt-Lunstad, Smith and Layton published this in PLoS Medicine in 2010, and it remains the reference point for the field. An odds ratio of 1.50 translates to roughly a 50 per cent greater likelihood of still being alive at follow-up, comparing people with stronger relationships against those with weaker ones.

Their comparison to other risk factors is what made the paper land. They concluded that the influence of social relationships on mortality sits alongside well-established risks including smoking and alcohol, and exceeds the effects of physical inactivity and obesity. Very few lifestyle findings carry that weight.

The second figure deserves attention too. Studies using complex measures of social integration, which look at the range and depth of a person's ties, produced a considerably larger estimate than studies asking a single question. How you measure connection changes what you find, and the better the measure, the bigger the effect looks.

Are loneliness and isolation the same thing?

No, and the distinction shows up in the numbers. A 2025 meta-analysis of 86 studies found social isolation carried the highest mortality risk at a hazard ratio of 1.35, living alone came next at 1.21, and loneliness, the subjective feeling, was lowest at 1.14.

Isolation describes your actual circumstances: how many people you see, how often, how embedded you are. Loneliness describes how it feels, which is the gap between the connection you have and the connection you want. A person can be isolated without feeling lonely, and can feel deeply lonely in a crowded life.

Nakou and colleagues, publishing in Aging Clinical and Experimental Research, found all three raise mortality risk, with objective isolation carrying more weight than the subjective feeling. This is worth sitting with, because the version most people worry about is the feeling, while the version the data flags hardest is the circumstance.

The practical reading is that the number and regularity of your actual contacts matters, and it matters somewhat independently of how content you feel with them.

Does this mean loneliness causes early death?

The evidence cannot establish that. Every study here is observational, following people over time rather than assigning them to be connected or isolated. Association across this many studies is meaningful, and it is a different claim from causation.

The obvious alternative explanation is reverse causation. People who are already unwell may withdraw socially, so poor health could be producing the isolation rather than the other way round. Good studies address this by excluding early deaths and adjusting for baseline health, and both meta-analyses report that the association survives those adjustments.

Nakou and colleagues also state their limitations plainly: substantial heterogeneity between studies, evidence of publication bias in some analyses, no standardised way of measuring loneliness across the field, and a heavy geographic concentration in Europe and North America. Holt-Lunstad and colleagues note the same regional skew and inconsistent statistical controls.

The grade on this page is evidence-based because the finding is large, replicated across many independent samples, consistent in direction, and supported by plausible biological pathways through stress physiology and inflammation. What it does not license is a promise that making a friend will extend your life by a specific amount.

What does this mean for how you live?

Treat your relationships as infrastructure worth maintaining, in the same category as sleep and movement. The data points toward regular contact with a range of people, and toward the practical circumstances of connection rather than only how connected you feel.

Two things follow from the isolation finding carrying more weight than the loneliness finding. Regular contact counts even on the days you do not especially feel like it, and structures that produce contact reliably, a standing weekly meeting, a class, a shared activity, do more than resolving to be more social.

The stronger effect for complex social integration measures suggests range matters alongside depth. A life with a few close ties and no wider circle scores differently from one with both.

This is also the area where the evidence is least ambiguous and most people invest least. Sleep and exercise get attention as health behaviours. On this data, who you see this week belongs in the same conversation.