Is a Polygenic Risk Score a 'Genetic Test'? Unresolved

The honest answer is that nobody has decided yet. A polygenic risk score sits in a gap the UK rules were not written to cover, and whether it counts as a "genetic test" for life insurance is genuinely unsettled — not settled and quietly hidden from you, but open, and flagged as open by the government's own review. So if you have read a confident line somewhere that an insurer can, or cannot, use a polygenic risk score, treat the confidence itself as a warning sign, and check the date the sentence was written.

What can be said firmly is narrower. Under the agreement that governs this whole subject in the UK — the Code on Genetic Testing and Insurance, a voluntary agreement between the government and the insurance industry — insurers do not today ask you for a polygenic risk score, and every commitment in that Code is built around predictive tests for specific inherited conditions. A polygenic score is a different kind of animal. It adds together the small effects of many common genetic variants into a single probability figure, rather than finding one fault tied to one disease. That difference is precisely why the classification is awkward, and why it has not been resolved.

The most recent three-year review of the Code, published in 2025, named this as a point that needs clarifying rather than one it was able to settle. So the position is a moving one, and this page is written in 2026 with that firmly in mind.

What a polygenic risk score actually measures

A single-gene predictive test asks a yes-or-no question about one fault. Do you carry the BRCA1 variant that runs in your family? Do you carry the expansion that causes Huntington's disease? The answer points at one condition, and the inheritance is usually straightforward enough that a clinician can talk about your risk in concrete terms.

A polygenic risk score does something structurally different. It scans hundreds or thousands of common variants, each of which nudges your risk of a condition up or down by a tiny amount, and it sums those nudges into one number. That number places you somewhere on a distribution — nearer the middle or nearer one tail — for something like coronary heart disease, type 2 diabetes or a common cancer. It is a statement about probability across a population, not a fault you either have or do not have.

Two things follow from that, and both matter for insurance. First, a polygenic score is rarely a verdict: a high score does not mean you will develop the condition, and a low one does not make you safe, because lifestyle, environment and plain chance still do most of the work. Second, the score is only as good as the reference data behind it, and much of that data was gathered in populations of European ancestry, so the scores perform unevenly across groups. A tool that fuzzy is a poor fit for a rulebook that was drafted around clean single-gene results.

Why the Code does not obviously cover it

The Code on Genetic Testing and Insurance is written around the idea of a predictive genetic test: a test carried out on someone without symptoms, looking for an inherited variant associated with a specific condition, in order to estimate future risk. That is the thing insurers have agreed not to ask for or use. If you want the history of how this agreement replaced the older voluntary arrangement, we cover it in what replaced the genetic testing moratorium in 2018.

Read that definition closely and a polygenic score does not slot neatly inside it. It is not aimed at one variant or one named condition; it is an aggregate. You could argue it is still a predictive genetic test, because it uses genetic data to predict future risk in a healthy person — which is the whole point of the protection. You could equally argue it is not the kind of test the Code was drafted to catch, because it does not identify a specific inherited fault at all, and behaves more like a statistical risk model dressed in genetic clothing. Both readings are defensible, which is another way of saying the text does not decide the question on its own.

This is not a hair-splitting distinction for lawyers only. The difference between a predictive and a diagnostic test is the hinge the whole system turns on, and it is worth being sure which side of it any test you have had actually falls. We work through that line with real cases in ten worked UK examples of predictive versus diagnostic tests. A polygenic score is the situation those examples do not cleanly resolve, because it is neither.

What the 2025 review said, and what it did not

The Code is reviewed roughly every three years, and the review published in 2025 is the document that matters here. On polygenic and integrated risk scores it did something unusual for an official document: it acknowledged the classification was unclear and set it down as work to be done, rather than pronouncing an answer. That is the single most useful fact on this page, because it tells you the uncertainty is real and officially recognised, not merely a gap in your own reading.

What the review did not do is change what happens on an application form today. Insurers are not asking for these scores, and there is no approved use of a polygenic result in the way Huntington's disease is the one approved use of a single-gene predictive result above a cover threshold. So the practical answer for now is stable even though the theoretical answer is not.

The one hard threshold in this whole area is worth stating plainly, because it is the only place a predictive genetic result can currently reach a UK life application at all.

Source: Code on Genetic Testing and Insurance, HM Government and the Association of British Insurers, published on gov.uk, together with its three-year review published in 2025.

Under that Code, the only predictive genetic result an insurer may ask about is a Huntington's disease test, and only once the total life cover applied for exceeds £500,000. Nothing else on the genetic side is askable, and a polygenic risk score is not among the exceptions. If your total cover is heading towards that figure for other reasons, the way separate policies can stack towards it is worth understanding in advance, which we set out in how a relevant life policy and personal cover can cross £500,000.

What insurers do in practice today

Set the theory aside and look at the form in front of you. A life insurance application asks about your own health, your measurements, your habits and, usually, your family history. It does not have a box for your polygenic risk score, and the insurer will not write to a research study to obtain one. What an underwriter reads is your answers to the questions asked, plus — with your consent — what your GP records show.

That last point is where a polygenic score could, in principle, become visible without anyone asking for it directly: if a score you obtained were written into your medical notes and led to monitoring, medication or a referral, the clinical care that followed would be ordinary medical history, and ordinary medical history is disclosable when the insurer asks about it. The score itself is contested ground; a statin prescription or a cardiology appointment is not. Before any of that reaches an underwriter you have a right to see what your doctor is about to send, which we explain in reading the GP report before it is sent.

It also matters when the score arrives relative to the policy. A result you obtain after cover is already in force does not reopen a policy that was honestly applied for; the position there is different again, and we cover it in having a genetic test after the policy started.

Research results, the NHS trials, and the exemption that does apply

Most polygenic scores that people encounter come out of research. The NHS and large UK studies have been trialling integrated risk tools that combine a polygenic score with ordinary clinical factors such as age, blood pressure and cholesterol, to see whether they sharpen the prediction of conditions like heart disease. If your score came from taking part in one of those studies, there is a clear rule that helps you.

Results generated in the course of research are not disclosable under the Code, and insurers have agreed not to ask for them. The reasoning is deliberate: if people feared that joining a study would count against them at insurance time, fewer would take part, and the research the health service depends on would suffer. So a polygenic score produced purely as a research output sits outside the insurance system by design.

The complication is what happens next. If a research finding prompts you to have an NHS test, or triggers clinical follow-up, that clinical information is ordinary medical history and does not carry the research exemption across with it. The protection attaches to the research result, not to everything that flows from it. In short: taking part is generally safe; acting clinically on what you learn produces records like any other patient's.

What this means if you are arranging cover now

If you are weighing up life insurance and you have, or are thinking about, a polygenic risk score, the sensible reading today is calm rather than fearful. The score is not a question on the form, it is not one of the results an insurer may ask about, and if it came from research it is expressly out of scope. Your application will still turn on the ordinary things — your age, your health, your weight, whether you smoke, and the family history questions insurers are entitled to ask at any level of cover regardless of any genetics rule.

What would be unwise is to treat the unsettled classification as a reason to rush, or to delay a clinically useful test, on the strength of a rule that might move. If a test would help your health, the insurance question should not be the thing that decides it. And because the answer here genuinely carries a date, it is worth reading around the wider subject before you act: our guide to family history, inherited risk and the odds pulls the threads together, and when you are ready to see real figures against your own circumstances you can get life insurance quotes and have an FCA-authorised broker work through them with you.

Scotland and Northern Ireland

The Code on Genetic Testing and Insurance is a single UK-wide agreement, so the insurance answer does not change as you cross a border: a polygenic risk score is no more askable in Edinburgh or Belfast than in Cardiff or Leeds, and the £500,000 Huntington's threshold is the same throughout. What differs between the nations is the machinery around it — how clinical genetics services are organised, how research studies recruit, and, for the separate right to see a GP's report, the fact that Northern Ireland relies on its own 1991 legislation rather than the Great Britain Act.

The other thing that varies by nation is the population data underneath any risk figure, which is why, if you are trying to benchmark your own family's risk, local numbers beat UK-wide averages. We show why that matters for one common condition in using Scottish numbers for a Scottish family's cancer history.

Frequently asked questions

Is a polygenic risk score a genetic test for insurance purposes?

It is not settled. The Code is written around results that identify variants associated with specific inherited conditions, whereas a polygenic score aggregates many common variants into a single probability instead. The 2025 three-year review of the Code flagged this point for clarification rather than resolving it, so any firm answer you read should be treated with caution and dated.

Could an insurer ask me for a polygenic risk score?

In practice insurers do not ask for these scores today, and the Code's commitments are framed around predictive tests for specific inherited conditions. Because the classification is unresolved, the safest reading is that the position may change over time, so any answer you rely on should carry the date it was written and be checked against the current version of the Code.

Should I avoid a research study that produces one?

There is no insurance reason to. Results generated in research are not disclosable under the Code, which is one reason taking part is generally safe. The complication comes afterwards: if a research finding leads to NHS testing or clinical follow-up, that clinical information is ordinary medical history and sits outside the research exemption, so it can be asked about like any other record.