Social Prescribing for Older Adults: What the Trials Actually Show

The United Kingdom built social prescribing into national policy in 2018, embedding it as one of the central mechanisms of the first government loneliness strategy anywhere in the world. Since then, the model — a clinician or link worker referring a patient to a community activity, class, or group rather than, or alongside, a medical treatment — has spread widely, including into programmes aimed specifically at older adults. The policy question is straightforward: does social prescribing reduce loneliness or isolation in this population, and does the trial evidence support the scale at which it has been adopted?

The honest answer is that it does not, not yet, in the form policymakers would need.

What the reviews actually contain

Two systematic reviews published in 2021 looked specifically at social prescribing and loneliness. One, in the International Journal of Environmental Research and Public Health, reported increases in self-esteem and self-confidence among participants, but noted limited trial evidence and substantial heterogeneity across the programmes it covered. The other, in Perspectives in Public Health, found that all nine included studies reported positive individual impacts, and three reported reductions in use of GP, emergency, social worker, or inpatient services. That sounds like a favourable verdict. It is not the same as a rigorous one.

Both reviews were built almost entirely from uncontrolled evaluations: pre-post designs, participant self-report, programmes assessing their own outcomes with no comparison group. A 2022 qualitative meta-synthesis in BMC Health Services Research adds texture rather than rigor: participants describe benefits extending beyond social contact itself, toward restored purpose and meaningful participation, and structured, purposeful group activity appears to work better than unstructured contact. That is a useful clue about mechanism. It is not evidence of effect size, because none of these designs can rule out the possibility that people who improve are simply the people who were already inclined to.

A 2025 systematic review protocol posted on medRxiv states the problem plainly: the effectiveness of social prescribing for older adults specifically remains unclear despite growing adoption, and the protocol’s authors could identify only one peer-reviewed randomised controlled trial in the entire area. One trial. Against a policy that has been written into national strategy for seven years and exported to health systems well beyond the UK.

The one trial that exists

That trial is worth naming, because it shows what the evidence base is missing. The HEAL-HOA study, published in The Lancet Healthy Longevity in November 2024, was a dual randomised controlled trial testing prosocial engagement and volunteering — a social-prescribing-adjacent intervention — against a control condition among lonely older adults in Hong Kong. Its value is structural: it is one of the only studies in this literature with a genuine control arm, allowing some claim to causal inference rather than correlation. That it took until late 2024 for a trial of this kind to appear, in a field with a national strategy dating to 2018, says something about how policy has run ahead of evidence here.

This is worth setting against the parallel epidemiological case for taking older-adult loneliness seriously. The National Academies concluded in 2020 that roughly one quarter of U.S. adults aged 65 and older are socially isolated, and called on the health care system to routinely assess isolation and loneliness — a recommendation echoed by clinician-facing commentary the same year urging routine screening in geriatric practice. AARP’s 2025 follow-up survey, directly comparable to its 2018 instrument, reports rising loneliness among adults 45 and older over that period. The scale of the problem is not in dispute. What remains unsettled is whether the specific remedy governments have adopted — referral into community activity via a link worker — actually moves the needle on it.

Why this gap matters for policy, not just for research

The distinction between isolation and loneliness sharpens the problem rather than resolving it. Social prescribing programmes typically enrol people because they are isolated — living alone, contact with few others, thin networks — and then measure success partly through self-reported loneliness, a subjective state that does not move in lockstep with objective network size. A programme can increase someone’s number of weekly social contacts without touching how lonely they feel, and the uncontrolled designs dominating this literature are poorly suited to catching that mismatch.

None of this means social prescribing does nothing. The consistent finding, across otherwise weak-design studies, that structured and purposeful activity outperforms unstructured contact is a real signal worth taking into intervention design, even if it cannot yet be attached to a number. But a policy built on nine uncontrolled evaluations and one randomised trial from a different health system than the one adopting it broadly is a policy running well ahead of its evidence.

What would close the gap is not more programme evaluation. It is more trials structured like HEAL-HOA: randomised, with genuine control arms, tracking isolation and loneliness as separate outcomes over a follow-up period long enough to distinguish a transient boost in contact from durable change. Health systems that have committed to social prescribing at scale have an obvious interest in producing that evidence themselves, since they are the ones already running the programmes. Until it exists, the honest framing for policymakers is that social prescribing is a plausible intervention with encouraging but methodologically weak support, not a proven one.