Kept.
← All research

Does loneliness actually affect your health?

Strong social relationships predict a 50% higher chance of survival

Holt-Lunstad, Smith & Layton (2010) · PLoS Medicine

148 studies, 308,849 participants, 7.5-year average follow-up. Association, not causation.

People with stronger social relationships had a 50% increased likelihood of survival over the follow-up period. The effect held across age, sex, initial health status and cause of death. It was strongest for complex measures of social integration and weakest for crude indicators like living alone.

How it was done

A meta-analysis of 148 studies covering 308,849 participants, with an average follow-up of 7.5 years.

What it means

The moderator is more interesting than the headline. The association is strongest for multidimensional measures of social integration and weakest for "how many people do you live with" — structure and quality matter more than headcount.

The authors note the magnitude is comparable to quitting smoking and exceeds obesity and physical inactivity, which is why this meta-analysis is cited so widely outside its field.

The WHO reached a similar conclusion in 2025, estimating that loneliness affects around one in six people worldwide and is linked to some 871,000 deaths a year.

A 2023 meta-analysis in Nature Human Behaviour (Wang et al., 90 cohort studies, 2.2 million people) updated the picture and split it: being socially isolated went with a 32% higher risk of death from any cause, feeling lonely with 14%. What people have counts for more in this literature than how they feel about it.

An older hand held in a younger one.
Kindel Media / Pexels

What it doesn’t show

Observational throughout — this is an association, and reverse causation is entirely plausible, since being unwell reduces social contact. It says nothing about whether changing someone’s social relationships changes their mortality, because no such trial exists. A 2024 genetic analysis of 476,100 people (Liang et al., Nature Human Behaviour) found that most of the observed links between loneliness and specific diseases did not hold up under Mendelian randomization, and suggested loneliness may be a marker of existing illness rather than its cause. So: linked to, never causes.

What Kept does about it

It is why the problem is worth working on, and nothing more than that. Kept makes no health claim of any kind. This study describes a population-level association; it does not describe what an app does to it.

Read the source

Holt-Lunstad, Smith & Layton (2010) — PLoS Medicine ↗

Confidence: A — meta-analysis, replicated finding, or official body. We don’t publish anything weaker than this.

Related

Kept is what we built from this.

A private voice journal for the people closest to you. Nine seconds of talking after you see someone, and the detail comes back before it matters.