Farnflower Newsletter
Connect

Is loneliness as harmful as smoking?

Where the famous comparison with cigarettes comes from, what its own authors say about it, and what happened to the risk in a study of half a million adults.

How connected do you feel? Take the one-minute test

Isolation, loneliness and early death

After accounting for health, habits and income, social isolation still went with a higher risk of early death, while loneliness alone didn’t.

Risk of dying during about 6.5 years (not isolated or not lonely = 100), basic vs full adjustment, 466,901 UK adults

  1. Isolated, basic adjustment 173
  2. Isolated, full adjustment 126
  3. Lonely, basic adjustment 138
  4. Lonely, full adjustment 99
Farnflower Source: Elovainio et al., 2017 · UK Biobank

Not in a simple, one-to-one way. The comparison comes from a large review that found weak social ties carried a risk of early death in the same range as smoking, and a US Surgeon General’s advisory later put it in cigarettes a day. Its own authors have since said the analogy oversimplifies, and in a large British study the feeling of loneliness no longer predicted death once health, habits and income were counted, while being socially isolated still did.

Key takeaways

  • The original review measured many kinds of social ties, not loneliness alone.
  • The strongest link was for being well integrated in social life, the weakest for simply living alone.
  • Smoking is the most thoroughly documented cause of disease, while the social findings come from observational studies.
  • Being socially isolated and feeling lonely are different things, and they behaved differently in the British data.
  • In statistical terms, mood, habits and money accounted for much of the extra risk of lonely people.

Where does the “15 cigarettes” line come from?

From a 2010 review, later turned into a cigarette count. In 2010, a team led from Brigham Young University pooled 148 studies with 308,849 people and found that people with stronger social relationships had a 50 percent greater likelihood of surviving during follow-up. They concluded that the influence of social relationships on the risk of death was comparable with that of smoking and alcohol and larger than that of physical inactivity and obesity, and showed this in a comparison chart (Holt-Lunstad et al., 2010).

Later documents, including the US Surgeon General’s 2023 advisory on loneliness and isolation, state it as a cigarette number. As a 2026 paper by the same research group summarizes it, lacking social connection can increase the risk of premature death comparable to smoking up to 15 cigarettes a day (Prochnow et al., 2026). Note the “up to”. In headlines it usually disappears.

What the 2010 review measured matters too. It pooled very different measures of social ties. The link was strongest for complex measures of how integrated people were in social life and weakest, and not statistically clear, for simply living alone (Holt-Lunstad et al., 2010). So the famous number isn’t really about the feeling of loneliness. It is about the whole web of people around you, which is the subject of Do friends help you live longer?

Is the comparison with smoking fair?

Partly, and its own authors say it oversimplifies. In a 2023 commentary, the lead author of the 2010 review and two colleagues wrote that the smoking comparison helped raise awareness of strong evidence, but that it often oversimplifies the evidence. They also warned that it may push attention toward treating lonely individuals rather than toward the structures and environments that make healthy relationships easier or harder (Smith et al., 2023).

I would add one plain difference. The 2010 review itself calls smoking the most extensively documented cause of disease ever investigated (Holt-Lunstad et al., 2010). The social findings come from observational studies, which are strong and consistent but can’t fully separate cause from effect, since being ill can also make people more isolated.

What happens to the risk when you count everything else?

It shrinks, and this is the part most articles leave out. A team led from the University of Helsinki analyzed 466,901 adults in the UK Biobank who answered questions about social isolation, meaning a score built from living alone, visits from others and social activities, and about loneliness, meaning feeling lonely and rarely having someone to confide in. You can check the feeling side yourself in the test How connected do you feel? They followed them for about six and a half years (Elovainio et al., 2017).

With only basic adjustments, for things like age and existing illness, both isolation and loneliness went with a higher risk of dying. After the researchers also accounted for income, education, smoking, drinking, exercise, depressive symptoms, blood pressure, body weight, grip strength, thinking skills and self-rated health, isolation still went with a higher risk, but loneliness no longer did (Elovainio et al., 2017).

That doesn’t mean loneliness is harmless. Many of the factors they adjusted for, such as depressive symptoms or smoking, may be part of how loneliness affects health, although a statistical analysis like this can’t prove the pathway. The authors draw the practical conclusion themselves: tackling poverty, unhealthy habits or poor mental wellbeing might reduce the extra deaths among isolated and lonely people (Elovainio et al., 2017). I find that more useful than the cigarette comparison, because it tells you where to start.

What I’d do

  • Take loneliness seriously as a reason to act, without panicking about a cigarette count.
  • Build one regular contact into your week, such as a fixed walk, class or call; a close relationship counts most when it is a good one, as in Do married people live longer?
  • Look after the things the study counted, such as your mood and how much you move.
  • Talk to your doctor if loneliness comes with low mood that lasts.
  • Get your hearing checked if conversations have become hard to follow, since hearing loss can shrink a social life, as Can hearing aids lower your risk of dementia? explains.

Frequently asked questions

How do I know if I’m lonely or just like being alone? Researchers measure loneliness as a feeling, not as time spent alone, so someone who enjoys solitude and has people to turn to isn’t lonely in their sense. The short questionnaire they use is the one in the test linked above.

What actually helps against loneliness? In a review of loneliness programs, the randomized trials that worked best addressed the negative expectations that keep lonely people from reaching out, rather than only creating chances to meet (Masi et al., 2011).

Sources

  • Elovainio, M., Hakulinen, C., Pulkki-Råback, L., Virtanen, M., Josefsson, K., Jokela, M., et al. (2017). Contribution of risk factors to excess mortality in isolated and lonely individuals: An analysis of data from the UK Biobank cohort study. The Lancet Public Health, 2(6), e260–e266. https://doi.org/10.1016/s2468-2667(17)30075-0
  • Holt-Lunstad, J., Smith, T. B., & Layton, J. B. (2010). Social relationships and mortality risk: A meta-analytic review. PLoS Medicine, 7(7), e1000316. https://doi.org/10.1371/journal.pmed.1000316
  • Masi, C. M., Chen, H. Y., Hawkley, L. C., & Cacioppo, J. T. (2011). A meta-analysis of interventions to reduce loneliness. Personality and Social Psychology Review, 15(3), 219–266. https://doi.org/10.1177/1088868310377394
  • Prochnow, T., Hawkley, L., Holt-Lunstad, J., Patterson, M. S., Merianos, A. L., & Smith, M. L. (2026). Clearing the air: Using tobacco prevention lessons for social connection. Public Health Reports, 141(2), 156–161. https://doi.org/10.1177/00333549251376268
  • Smith, R. W., Holt-Lunstad, J., & Kawachi, I. (2023). Benchmarking social isolation, loneliness, and smoking: Challenges and opportunities for public health. American Journal of Epidemiology, 192(8), 1238–1242. https://doi.org/10.1093/aje/kwad121

One question a month, answered properly