Half a million people just joined a study with no end date
Our Future Health enrolled 500,000 UK adults for lifelong data collection, funded by government and drugmakers jointly.
By The Weekend · · 5 min read
Our Future Health has recruited more than 500,000 adults in the UK, making it the largest health research cohort the country has assembled, according to the program's own account in Nature Medicine. Participants gave blood, filled out questionnaires, and agreed to let researchers watch their health records for years, maybe decades. No one has said when, or if, that watching stops.
A biobank is not a study. It is infrastructure for studies that don't exist yet.
Most clinical trials ask a narrow question: does this drug help this condition in these patients, measured over this many months. Our Future Health asks nothing specific. It collects genetic data, blood chemistry, lifestyle surveys and long-term health outcomes from half a million people, then makes that dataset available to researchers who haven't been recruited yet, for questions that haven't been written yet.
That's the model UK Biobank pioneered starting in 2006 with 500,000 participants recruited over four years. Our Future Health is a second, larger instrument built on the same premise: population-scale data outlives any single hypothesis. The bet is that pooling genetics with health records at this scale will surface disease patterns too rare or too slow-moving for a normal trial to catch — the kind of signal that needs a million data points, not a thousand.
The government funds recruitment. Drug companies fund what happens after.
Our Future Health is a charity, but its money comes from a mixed roster: UK government funding sits alongside contributions from pharmaceutical companies including AstraZeneca, GSK and others with a direct commercial interest in what the dataset finds. The structure isn't hidden — it's listed as a public-private partnership — but it does mean the same companies that might one day sell a drug based on a pattern found in this data helped pay to find the pattern.
That's not disqualifying. UK Biobank ran on a similar mix and produced research that held up under scrutiny, including large genetic studies published in Nature and other peer-reviewed venues over the past decade. But it does mean the usual question — who funded this, and what did they gain — has a two-part answer here. The government gains a public health tool and, eventually, policy-relevant findings. The drug companies gain a shorter path to the population subgroups their next trial needs to find, recruit and prove a drug works in.
What the researcher gets: a shortcut past the hardest part of any trial
The expensive, slow part of a clinical trial is rarely running it — it's finding the right participants. A trial testing a drug for a genetic subtype of a disease needs to locate people with that subtype, confirm eligibility, and recruit them one by one, a process that can take years and frequently fails to hit enrollment targets at all.
A cohort like this flips that sequence. The genetic and health data already exist, collected once, sitting in a database. A researcher designing a trial for, say, a rare cardiovascular variant can query the dataset, find several thousand matching participants, and approach only them — turning a multi-year recruitment slog into a database search. UK Biobank has enabled more than 30,000 published research projects using exactly this shortcut since it opened its data to approved researchers.
That's the mechanism worth understanding: this isn't primarily a study of half a million people's health. It's a matching engine that lets future studies skip their hardest step. The value compounds with time — the longer participants are followed, the more disease onset, treatment response and long-term outcome data accumulates, and the more valuable the dataset becomes to researchers who had nothing to do with building it.
What participants signed up for isn't fully specified, and that's the design
Consent for a normal trial covers a defined intervention and a defined endpoint. Consent for a biobank covers an open-ended commitment: your data, available to researchers vetted by a governance committee, for studies that haven't been conceived yet, indefinitely. UK Biobank's own consent framework works this way, and Our Future Health follows the same logic — participants can withdraw, but the studies already run using their data don't get unwound.
This isn't a criticism unique to Our Future Health; it's the structural trade-off every population biobank asks participants to accept, and regulators have generally accepted it as ethical provided access is governed, anonymized and reviewed case by case. But it's worth naming plainly rather than letting "research partnership" language smooth over what's actually being asked: an indefinite data-sharing commitment in exchange for no direct benefit to the person giving the data, only to whoever studies it later.
The number that will decide if this was worth building
UK Biobank's real test came roughly a decade after launch, when its genetic data started producing findings that changed clinical guidance — including risk scores for cardiovascular disease now used in some NHS screening tools. Our Future Health, launched with a similar half-million-person target, is at year one of what will likely be a similar wait.
The falsifiable version of the bet: if Our Future Health is working, expect the first wave of peer-reviewed findings — genetic risk scores, treatment-response patterns, subgroup discoveries — to start appearing in major journals within five to eight years, roughly the timeline UK Biobank took to produce its first clinically actionable results. If nothing publishable emerges by then, the dataset will have cost its funders a decade of collection and governance overhead for a return that never materialized. Right now, the honest answer to what this has accomplished is: it has assembled the raw material. The research that justifies assembling it hasn't been written yet.