What Reduces Loneliness? What the Evidence Shows

Loneliness has become a public health issue. In 2018 the United Kingdom appointed a minister for loneliness. In 2023 the US Surgeon General issued an advisory on what he called an epidemic of loneliness and isolation. In 2025 the World Health Organization’s Commission on Social Connection estimated that about one person in six worldwide is lonely and linked loneliness to around 871,000 deaths a year.

The obvious response is to bring people together: lunch clubs, befriending calls, community groups, apps that match neighbours. Governments and charities have funded thousands of such schemes. The question this article asks is simpler and harder: which of them actually reduce loneliness, by how much, and for how long?

The research has grown fast. A meta-analysis published in 2026 combined 280 studies. The main findings are encouraging in one way and sobering in another. Interventions do reduce loneliness on average, and the ones that work best are not the ones most people would guess. But the confidence in almost every estimate is low, the effects shrink in the most rigorous studies, and there is very little evidence about what happens after the programme ends. We looked at the link between loneliness and mortality in a separate article.

Loneliness is not the same as being alone

Researchers distinguish two things that everyday language often mixes up. Social isolation is objective: how many people you see, how often, and how many you could call on. Loneliness is subjective: the distress of feeling that your relationships are fewer or less close than you want. People can be isolated without feeling lonely, and lonely in a crowd or a marriage.

The distinction matters for interventions. A scheme that increases the number of contacts may reduce isolation without touching loneliness, if the contacts are not the kind the person wants. That is one reason why simply arranging more social contact has turned out to be less effective than expected.

Why loneliness can feed itself

The late John Cacioppo, one of the founders of loneliness research, and his colleague Louise Hawkley proposed an explanation for why loneliness is hard to shake. Feeling isolated puts the brain on alert for social threat. Lonely people become more sensitive to signs of rejection, remember negative encounters more readily, and expect others to be less friendly. They may then behave more cautiously or defensively, which can make encounters colder and confirm the original fear.

This model, set out in a 2009 review in Trends in Cognitive Sciences, explains why the most effective interventions target thinking rather than contact alone. It also explains a common frustration: telling a lonely person to ‘just get out more’ often fails, because the problem is not simply a lack of opportunities.

Loneliness is also often triggered by life changes that remove relationships all at once: bereavement, divorce, retirement, moving to a new city, becoming a carer or developing a disability. Periods like these are when support is most likely to help, and when people are least likely to ask for it.

The first big synthesis

In 2010, Christopher Masi, Hsi-Yuan Chen, Louise Hawkley and John Cacioppo published a meta-analysis of loneliness interventions in Personality and Social Psychology Review. They grouped programmes into four strategies: improving social skills, enhancing social support, increasing opportunities for social contact, and addressing maladaptive social cognition, the habits of thought that lead lonely people to expect rejection and read neutral situations as hostile.

Two findings from that paper still shape the field. First, the study design mattered enormously. Single-group before-and-after studies and non-randomised comparisons produced much larger effects than randomised trials. Second, among the randomised trials, the most successful interventions were those that addressed maladaptive social cognition, usually with techniques drawn from cognitive behavioural therapy.

What the first major meta-analysis found

Interventions to reduce loneliness, grouped by study design and strategy.

Question	Finding
Which designs showed the biggest effects?	Single-group and non-randomised studies
Which designs showed the smallest?	Randomised trials
Which strategy worked best in randomised trials?	Addressing maladaptive social cognition

Masi CM, Chen HY, Hawkley LC, Cacioppo JT, Personality and Social Psychology Review, 2011;15(3):219–266.

What 280 studies show

The largest synthesis so far, by Mathias Lasgaard, Pamela Qualter and colleagues, was published in the American Psychologist in 2026. It combined 280 studies of interventions for people of all ages. In 122 randomised trials, the average short-term effect, measured up to four weeks after the programme, was a standardised mean difference of −0.50, a small to moderate reduction in loneliness.

The authors then rated their confidence in that estimate using the GRADE system, which considers bias, consistency and precision. It was low or very low. They found no significant differences between age groups. Psychological interventions appeared to be the most effective, with moderate effects, while social and emotional skills training, social network interventions and social support showed smaller ones.

Loneliness interventions, 280 studies

Short-term effect, up to four weeks after the programme, in randomised trials.

−0.50average effect on loneliness in 122 randomised trials (standardised)

LowGRADE confidence in the estimate: low or very low

Lasgaard M, Qualter P, Løvschall C, and colleagues, American Psychologist, 2026;81(1):36–52.

Other reviews point the same way. A 2021 meta-analysis of psychological interventions by Nadine Hickin, Anton Käll and Roz Shafran in Clinical Psychology Review found an average effect of 0.43. A 2025 rapid review of 101 interventions by Joanna Blodgett and colleagues, in the Journal of Public Health Policy, ranked psychological interventions first, with an effect of 0.79, followed by interventions based on social interaction at 0.50 and on social support at 0.34.

Which kinds of programme reduce loneliness most

Pooled standardised effect sizes, 101 interventions.

Psychological interventions0.79

Social interaction0.50

Social support0.34

Blodgett JM, Tiley K, Harkness F, and colleagues, Journal of Public Health Policy, 2025;46(2):245–268. Includes non-randomised studies.

The phone call that worked, and what it cannot tell us

One of the best-known recent trials tested something simple. During the COVID-19 pandemic, Maninder Kahlon and colleagues at the University of Texas randomised adults receiving home-delivered meals to either regular phone calls from trained lay volunteers or no calls. The callers were taught to listen with empathy and to let the participant guide the conversation. The results, published in JAMA Psychiatry in 2021, showed a meaningful drop in loneliness after four weeks, along with reductions in depression and anxiety.

The authors were candid about the limits. They could not tell whether the benefit came from being called at all or from the empathetic style of the calls, since the control group received nothing. The trial measured outcomes only four weeks after it started, so it could not say whether the effect lasted. And 13 people dropped out of the call group against one in the control group, some after only two calls, citing time and interest.

A larger trial in Hong Kong addressed some of these problems. Published in Age and Ageing in 2025 by Da Jiang and colleagues, it compared telephone-based behavioural activation and mindfulness, delivered by lay counsellors, with ordinary befriending calls as an active control, among older adults who lived alone and lacked digital access. Both structured programmes reduced loneliness more than befriending, and the authors reported outcomes at six months. The comparison is a reminder that contact alone is not the active ingredient; what is done with the contact matters.

Empathetic phone calls, randomised trial

Adults receiving home-delivered meals, four weeks of calls versus no calls.

0.48effect size on loneliness after four weeks

13 vs 1dropouts in the call group and the control group

Kahlon MK, Aksan N, Aubrey R, and colleagues, JAMA Psychiatry, 2021;78(6):616–622; effect size as reported by Tang VFY et al, 2025.

Why the effects shrink

Several features of the research explain why confidence is low. Many studies are small and short. Many measure loneliness only immediately after the programme, when participants may still feel the warmth of the group or want to please the researchers. The Lasgaard review found far fewer studies of long-term effects than of short-term ones. And most programmes are tested by the organisations that designed them, which can influence how results are measured and reported.

There is also a subtler problem. People who sign up for a loneliness programme are often at a low point, and loneliness tends to fluctuate. Some improvement would happen without any help, a pattern known as regression to the mean. That is one reason why single-group studies, which cannot account for it, have always shown bigger effects than randomised trials.

Apps, video calls and robots

Digital interventions have multiplied since the pandemic, and they are now tested in randomised trials. A 2025 meta-analysis by Thomas Hansen and colleagues in Internet Interventions pooled 40 trials with 6,062 participants, a third of them focused on younger people. Psychological programmes delivered online, especially those with a group or social element, reduced loneliness. So did group-based activities and, perhaps surprisingly, robotic pets.

Other digital approaches did not show clear benefits. Interventions that simply increased contact, self-guided activities done alone, and conversational robots designed to chat with users were not effective in the pooled data. As with face-to-face programmes, the pattern suggests that what helps is changing how people relate to others, not just adding more interactions.

Reviews of loneliness interventions

Main systematic reviews and meta-analyses, in order of publication.

Review	Scope	Main finding
Masi and colleagues, 2011	All ages	Randomised trials show smaller effects; changing social cognition works best
Hickin and colleagues, 2021	Psychological interventions	Average effect 0.43
Blodgett and colleagues, 2025	101 interventions	Psychological 0.79, social interaction 0.50, support 0.34
Hansen and colleagues, 2025	40 digital trials	Online therapy, groups and robotic pets helped
Lasgaard and colleagues, 2026	280 studies	Effect −0.50 short term; confidence low

Sources as listed at the end of this article.

Social prescribing

One of the most widely adopted responses is social prescribing, in which doctors refer patients to a link worker who connects them with local activities such as walking groups, arts classes or volunteering. England made it part of its national health service plans in 2019, funding link workers in general practice, and many countries have followed with their own versions.

Reviews of social prescribing have generally found that participants like it and report feeling better, but that the evidence on loneliness specifically is weak, based largely on uncontrolled studies and qualitative interviews. That does not mean it fails, and some people clearly benefit. It means that a policy has spread faster than the evidence that could show whether, and for whom, it works.

A 2024 review for older adults living in the community, led by Paul Shekelle and published in the Journal of General Internal Medicine, reached a similar judgement. Of 60 studies it included, only a minority provided evidence it considered of moderate certainty, and effects that were found tended to be modest.

What this means for individuals

The evidence points to a few practical conclusions. First, loneliness is partly a matter of how we interpret social situations, and that can be changed. Approaches based on cognitive behavioural therapy, which help people notice and question assumptions such as ‘they didn’t reply because they don’t like me’, have the strongest evidence.

Second, quality matters more than quantity. Regular contact with a purpose, such as volunteering, a class or a shared hobby, gives relationships time to develop, and structured, empathetic conversation appears more helpful than a quick check-in. Third, persistent loneliness often travels with depression and anxiety. If it is affecting sleep, mood or daily life, it is worth talking to a doctor; treating the depression may ease the loneliness, and the reverse. And if you are the friend or relative of someone who seems lonely, a regular, unhurried conversation in which you mostly listen may matter more than a busy social calendar arranged on their behalf.

Questions people ask

What is the most effective way to reduce loneliness?

Psychological approaches, especially those that address how people interpret social situations, have shown the largest effects in reviews.

Do befriending phone calls help?

A short course of empathetic calls reduced loneliness after four weeks in one US trial. In a Hong Kong trial, structured programmes did better than befriending calls.

Is loneliness the same as social isolation?

No. Isolation is having few contacts; loneliness is feeling that your relationships fall short. They overlap but need different responses.

How strong is the evidence?

Weak overall. A review of 280 studies found a small to moderate average effect, but rated confidence as low or very low.

Can apps or robots reduce loneliness?

Online psychological programmes and group activities helped in trials, as did robotic pets. Conversational robots and simple contact apps did not show clear benefits.

Does social prescribing work for loneliness?

Participants tend to value it, but controlled evidence on loneliness is limited.

The short version

• Interventions reduce loneliness on average, with a small to moderate short-term effect across 280 studies.
• Confidence is low: most studies are small, short and uncontrolled, and long-term data are scarce.
• Psychological approaches that address how people read social situations work best; arranged contact alone works least.
• A trial of empathetic phone calls cut loneliness in four weeks, but could not show the benefit lasted.
• Loneliness is not isolation: more contacts do not guarantee feeling less lonely.

This article summarises published research on loneliness interventions. It is not medical advice or a clinical assessment. If loneliness is affecting your mood, sleep or daily life, speak to a doctor or a qualified mental health professional.

Further reading. Lasgaard and colleagues, ‘Are loneliness interventions effective for reducing loneliness?’, American Psychologist, 2026, is the most comprehensive review. Masi and colleagues, 2011, remains the classic starting point.

Sources

1. Cacioppo JT, Hawkley LC. Perceived social isolation and cognition. Trends in Cognitive Sciences, 2009;13(10):447–454. doi:10.1016/j.tics.2009.06.005.
2. Masi CM, Chen HY, Hawkley LC, Cacioppo JT. A meta-analysis of interventions to reduce loneliness. Personality and Social Psychology Review, 2011;15(3):219–266. doi:10.1177/1088868310377394.
3. Lasgaard M, Qualter P, Løvschall C, and colleagues. Are loneliness interventions effective for reducing loneliness? A meta-analytic review of 280 studies. American Psychologist, 2026;81(1):36–52. doi:10.1037/amp0001578.
4. Hickin N, Käll A, Shafran R, and colleagues. The effectiveness of psychological interventions for loneliness: a systematic review and meta-analysis. Clinical Psychology Review, 2021;88:102066. doi:10.1016/j.cpr.2021.102066.
5. Blodgett JM, Tiley K, Harkness F, and colleagues. What works to reduce loneliness: a rapid systematic review of 101 interventions. Journal of Public Health Policy, 2025;46(2):245–268. doi:10.1057/s41271-025-00561-1.
6. Kahlon MK, Aksan N, Aubrey R, and colleagues. Effect of layperson-delivered, empathy-focused program of telephone calls on loneliness, depression, and anxiety among adults during the COVID-19 pandemic. JAMA Psychiatry, 2021;78(6):616–622. doi:10.1001/jamapsychiatry.2021.0113.
7. Jiang D, Kwok JYY, Yeung DY, and colleagues. Six-month outcomes of layperson-delivered, telephone-based behavioural activation and mindfulness interventions on loneliness among older adults. Age and Ageing, 2025;54(8):afaf209. doi:10.1093/ageing/afaf209.
8. Tang VFY, Jiang D, Kahlon M, and colleagues. A three-arm randomized controlled trial protocol of telephone-based interventions on loneliness in older adults. Contemporary Clinical Trials Communications, 2025. doi:10.1016/j.conctc.2025.101479.
9. Shekelle PG, Miake-Lye IM, Begashaw MM, and colleagues. Interventions to reduce loneliness in community-living older adults: a systematic review and meta-analysis. Journal of General Internal Medicine, 2024;39(6):1015–1028. doi:10.1007/s11606-023-08517-5.
10. Hansen T, Johansen RF, Kirkøen B, and colleagues. Digital bridges to social connection: a systematic review and meta-analysis of digital interventions for loneliness and social isolation. Internet Interventions, 2025. doi:10.1016/j.invent.2025.100856.
11. World Health Organization. From Loneliness to Social Connection: Charting a Path to Healthier Societies. Report of the WHO Commission on Social Connection, 2025.
12. Office of the US Surgeon General. Our Epidemic of Loneliness and Isolation. 2023.