Social Prescribing Works for Older Adults, Probably. Does It Work for Young Adults?

Social prescribing is the policy the United Kingdom built its loneliness strategy around in 2018, and it has since become the default tool that health systems reach for when loneliness is identified as a problem worth a budget line. A general practitioner, or a link worker attached to a clinic, refers a patient to a non-clinical activity — a walking group, a choir, a gardening project, a volunteering scheme — on the theory that structured social contact will reduce isolation and improve wellbeing. The U.S. Surgeon General’s 2023 advisory cites social connection as comparable in health impact to smoking or obesity, and social prescribing is one of the few concrete mechanisms on offer for acting on that finding.

The question this piece asks is narrower than “does social prescribing work.” It is: does the evidence support extending it, as currently designed and tested, to young adults specifically. The answer requires looking at who the existing trials actually enrolled.

What the systematic reviews establish, and about whom

Two systematic reviews published in 2021 looked at social prescribing’s effect on loneliness. One found that all nine included studies reported positive individual impacts, with three showing reductions in use of GP, emergency, social worker, or inpatient services. A companion review reported gains in self-esteem and self-confidence as a consistent outcome. A 2022 qualitative meta-synthesis added a specific mechanism: participants described benefit that went beyond social contact itself, toward restored purpose and meaningful participation, and structured, purposeful activity appeared to work better than unstructured contact alone.

That is a reasonably encouraging body of evidence. It is also, on inspection, a body of evidence about a different population than the one this article is asking about. The programmes underlying these reviews are drawn overwhelmingly from primary care referral pathways — patients already in contact with a GP, disproportionately older, often referred following a health event or a diagnosis of depression. None of the three reviews cited here report age-stratified results for adults under 30. The 2021 wellbeing review also flags the underlying problem directly: trial evidence is limited and the programmes evaluated are heterogeneous, which makes generalizing to any population, let alone a specific age band, a stretch.

The one genuine randomised controlled trial in this space — as opposed to an uncontrolled programme evaluation, which is what most of the social prescribing literature amounts to — is the HEAL-HOA study published in November 2024. It tested prosocial engagement and volunteering against a control condition among lonely older adults in Hong Kong. It is valuable precisely because RCTs are rare here; the rest of the field is dominated by small, unblinded, self-selected samples. But it is again a study of older adults. There is no equivalent RCT in the source record testing social prescribing, or anything like it, on adults in their late teens or twenties.

What is known about young adults’ loneliness, separately

The population that would receive this intervention, if it were extended downward, looks different from the ones it was tested on. The Harvard Graduate School of Education’s 2021 survey found 61% of young adults aged 18 to 25 reporting serious loneliness, against 36% of Americans overall — the highest rate of any group in that survey, and more than double the general figure. About half of lonely young adults said no one had taken more than a few minutes recently to ask how they were doing in a way that felt genuine. Cigna’s 2020 workplace survey found 73% of workers aged 18 to 22 reporting loneliness, and over 80% of employed Gen Z workers. The American Enterprise Institute’s 2021 survey on friendship found the share of Americans with no close friends had quadrupled since 1990, with the sharpest declines concentrated among men, a group that is also the least likely to be in regular contact with a GP or any other referral gatekeeper that a prescribing pathway depends on.

This matters for the policy question specifically because social prescribing, as designed, routes through primary care. A young adult who is lonely but not visibly unwell, not depressed enough to be flagged, and not seeing a doctor with any regularity has no obvious entry point into the referral pathway that the existing evidence was built to test. The intervention and the population most in need of it are not well matched by design, independent of whether the intervention itself works.

The one study that looks at young adults directly

The closest thing in the source record to intervention evidence for this age group is a 2025 study in Frontiers in Psychology examining which forms of social connection actually predicted wellbeing among college students coming out of the pandemic. It is not a trial of a policy intervention; it is an observational study of what kinds of contact correlated with better outcomes. But its finding is consequential for policy design regardless: quality and regularity of contact predicted wellbeing more strongly than sheer quantity of social ties. That finding, if it generalizes, cuts against a referral-and-attend model that measures success by attendance at a single weekly activity, and toward something that sustains contact over time. It is one study, cross-sectional in the relevant sense of measuring existing patterns rather than assigning an intervention, and it should be weighted accordingly.

What this adds up to

The evidence supports a claim considerably narrower than “social prescribing reduces loneliness in young adults.” What it supports is: social prescribing has plausible, moderately encouraging observational and qualitative support among older adults referred through primary care, backed by exactly one randomised trial, also in older adults. Extending that finding to people in their twenties requires two unproven assumptions — that the mechanism (structured group activity, referred through a clinical pathway) works the same way in a population with different referral access, and that the outcome (reduced loneliness, better wellbeing) responds to the same dose of contact. Neither assumption has been tested. The claim is not that the policy would fail; it is that no one has looked.

A study that would settle it would randomise young adults, recruited outside primary care — through universities, employers, or community organizations rather than GPs — to a structured, purposeful group activity against a passive control, and would measure loneliness with a standard instrument like the UCLA Loneliness Scale before and after, not just self-reported satisfaction with the programme. Until something resembling that exists, the honest description of current policy is that it is extrapolating from a different population, not applying a tested one.