Huntington’s Genetics and the 50 Per Cent: CAG Repeats Explained

Huntington’s genetics and the 50 per cent: CAG repeats explained

Huntington’s disease is caused by one fault in one gene, and the fault is a stutter.

On chromosome 4 sits a gene called HTT, which makes a protein called huntingtin. Within that gene is a short sequence — the DNA letters C, A and G — that repeats. Everybody has it. The question is how many times it repeats.

Below a certain number, nothing happens. Above it, the protein is made with an abnormally long tail, misfolds, accumulates, and kills neurons, starting in the striatum. The length of the repeat determines whether a person will develop Huntington’s disease, and influences roughly when.

This is unusual in medicine. Most genetic risk is probabilistic and multifactorial. Huntington’s is close to deterministic: above 40 repeats, a person will develop the disease if they live long enough. That is the fact that makes the genetics of HD so consequential for families, and it is why this guide exists before the one on testing.

It covers the four repeat ranges and what each means, the grey zone in between, how the number changes between generations, why it matters which parent passed it on, what the 50 per cent figure does and does not predict, and the reproductive options available in Australia.

A note before starting: nothing in this guide is a substitute for genetic counselling. Interpreting a repeat number for an actual family, with actual relatives and actual decisions attached, is specialist work. In Australia it is available free through state genetic services, and no number should be interpreted without it.

The four ranges

The Centre for Genetics Education, part of NSW Health, sets out the ranges used in Australian practice.

Four observations about that table that matter more than the numbers themselves.

The threshold is sharp but not absolute. Thirty-five repeats and thirty-six repeats are one DNA triplet apart and sit either side of a clinical boundary. That is uncomfortable and it is real. Somebody told they have 36 repeats has a genuinely uncertain future; somebody told they have 35 does not have the disease but may have children who do.

Reduced penetrance means what it says. Between 36 and 39, the gene may never cause symptoms, or may cause them at 75 instead of 45, or may cause a milder course. This range is small and it is the one where counselling matters most, because the honest answer is a range of possibilities rather than a prediction.

The intermediate range is how new cases appear. A person with 27 to 35 repeats will not develop HD and may have no family history at all. If the repeat expands when passed to a child, that child can land in the disease range — and the family’s first case of Huntington’s appears with no inherited history behind it. Roughly this mechanism accounts for the apparently sporadic cases.

Above 40, repeat length influences timing but does not fix it. Longer repeats are associated with earlier onset, and the relationship is strong at the extremes — very long repeats produce juvenile onset — but in the common range of 40 to 50 the correlation is loose enough that no useful individual prediction can be made from it. This is the single most important thing to understand about a repeat result, and it is covered below.

Why the number does not tell you when

People who receive a positive predictive test result almost always want the next number: when.

The repeat length will not give it to them. The association between repeat length and age at onset accounts for a substantial share of the variation at the population level and a frustratingly small share for any individual. Two people with 43 repeats can have onset fifteen years apart. Modifier genes elsewhere in the genome, somatic instability of the repeat in brain tissue over a lifetime, and environmental factors all contribute, and none of them are measurable in clinic.

The Centre for Genetics Education states it plainly: the test cannot predict when symptoms will start or how severe they will be.

What this means practically is that a positive result tells you the outcome and not the schedule. That is genuinely hard to live with, and it is also the reason the decision to test is a decision about information rather than about planning. See predictive testing for Huntington’s.

There is one partial exception. Where a very long repeat is found — above about 55 — juvenile onset becomes likely, and the timing becomes considerably more predictable. See juvenile Huntington’s disease.

Anticipation, and why it matters which parent

The repeat is not perfectly copied when it is passed on. It can expand or contract between generations, and the expansion is not symmetrical between mothers and fathers.

Fathers are far more likely than mothers to pass on an expanded repeat. The explanation generally offered is that sperm are produced continuously throughout a man’s life, with many more rounds of DNA copying than the eggs a woman is born with, and each copy is an opportunity for the repeat to lengthen. Older fathers appear more likely to transmit an expansion than younger ones. The child’s own sex makes no difference to any of this.

The clinical consequence is called anticipation: the disease appearing earlier and sometimes more severely in each successive generation. A father with 44 repeats and onset at 50 may have a child with 56 repeats and onset at 20.

Which is why juvenile Huntington’s disease is usually paternally transmitted. A family with juvenile-onset HD in a child almost always has a father, rather than a mother, carrying the gene — and sometimes a father who is not yet symptomatic himself.

Full penetrance repeats are the most unstable. Repeats of 40 or more are more likely to expand on transmission than intermediate or reduced penetrance ones, which are relatively stable but can still change. For intermediate and reduced penetrance alleles, the risk of expansion cannot currently be quantified, and is thought to be small — a genuinely unsatisfying answer for the families it affects, and the right one.

What the 50 per cent actually means

The figure gets misunderstood in several directions, so it is worth being precise.

It is 50 per cent per child, per pregnancy. A person with a full penetrance gene copy has one affected copy and one normal copy, and passes one of them to each child with equal probability. Having one child who inherited it does not change the odds for the next. Nor does having three who did not.

It is 50 per cent of inheriting the gene, which above 40 repeats is effectively 50 per cent of developing the disease. In the reduced penetrance range it is 50 per cent of inheriting a gene copy that may or may not cause symptoms.

Nothing about family resemblance or symptom patterns affects it. Whether a child looks like the affected parent, has their temperament, or was the one who always got the family illnesses is irrelevant. This sounds obvious written down and it is a near-universal source of informal reasoning in affected families.

It applies to siblings too. Each sibling of an affected person had the same 50 per cent at conception. A negative predictive test in one sibling tells the others nothing.

And a negative test genuinely is negative. This is the point most worth emphasising, because at-risk people frequently do not believe it. If the normal copy was inherited, the person will not develop Huntington’s disease and cannot pass it to their children. The 50 per cent is resolved, permanently, in that direction. About half the people who test receive that result.

One important asymmetry. Testing a person at risk also reveals information about their parents and siblings whether they want it or not. A positive result in somebody whose affected parent is unknown identifies which side of the family it came from. A positive result in a person with an untested parent effectively diagnoses that parent as a carrier. This is why genetic counselling spends so much time on who else will learn what, and it is why results are not released without written consent except where a court requires it.

If there is no family history

A substantial minority of people diagnosed with Huntington’s have no known affected relative. There are several explanations, and they are worth knowing because “it can’t be Huntington’s, there’s no family history” delays diagnoses.

A parent died before onset. Mean onset is around 45 and average age at death from HD is 54 to 58. A grandparent who died at 40 of something else may have been a carrier.

A parent was never diagnosed. Misdiagnosis as psychiatric illness or alcoholism is common, particularly in earlier decades.

Adoption, donor conception, non-paternity, or lost contact with one side of a family.

A new expansion from an intermediate allele, as described above — a parent with 27 to 35 repeats who will never develop HD passing on an expanded copy.

Or the family knew and did not say. Huntington’s carries enormous historical stigma and many families concealed it, sometimes across generations. Adults frequently discover at 40 that their grandmother’s “nervous condition” or their uncle’s “drinking problem” was this.

Testing a child

The position in Australian and international practice is consistent: predictive testing is for adults.

The Centre for Genetics Education states that eligibility for predictive testing begins at age 18. The reasoning is that a child cannot consent to receiving information about their own adult health, that there is no treatment that early knowledge would enable, and that the child’s right not to know belongs to them and should be exercised when they are able to exercise it. Published international recommendations for predictive testing in HD take the same position.

The exception is diagnostic testing. Where a child or adolescent has symptoms consistent with juvenile Huntington’s disease, a genetic test is appropriate because it is diagnosing a present illness rather than predicting a future one. That is a different test with a different purpose.

This is a painful rule for parents who want to know, and it holds up well in practice. Families who want help discussing it with children — and families whose teenagers are asking — can get it through genetic counselling and through Huntington’s Australia on 1800 244 735. See families and children.

Reproductive options in Australia

For somebody who carries the gene or is at risk and wants children, there are several routes. All of them warrant a genetics appointment before a pregnancy rather than during one.

Preimplantation genetic testing (PGT), formerly called PGD. IVF is used to create embryos, which are tested, and an unaffected embryo is transferred. The main advantage is avoiding a decision about an established pregnancy. The Centre for Genetics Education notes it is available in Australia through private fertility clinics, which means cost is a real factor, though Medicare rebates apply to parts of IVF.

Exclusion testing, a variant worth knowing about. For an at-risk person who does not want to know their own status, embryos can be tested to determine which grandparental chromosome 4 they inherited, and only embryos carrying the unaffected grandparent’s copy are transferred. This allows reproductive choice without revealing the parent’s own result. It is technically available and emotionally complex, and it needs specialist counselling.

Prenatal diagnosis. Chorionic villus sampling at around 11 to 12 weeks, or amniocentesis at around 15 to 19 weeks, both with a miscarriage risk under 1 per cent in experienced hands. The Centre for Genetics Education advises discussing this with a genetic counselling team before pregnancy, and having the at-risk parent tested first if time allows — because prenatal testing on a parent of unknown status creates a situation where a positive fetal result also reveals the parent’s own status.

Donor gametes.

Adoption.

Having children without testing, which is a legitimate choice that a great many people make.

And deciding not to have children, equally legitimate, and worth saying out loud because at-risk people sometimes feel they have been handed a decision rather than making one.

What matters here is that all of these are genuine options and several of them are poorly known. The reproductive counselling appointment is one of the highest-value hours available to an at-risk person, and it is free through public genetic services.

Insurance, briefly — because the law changed

This is covered properly in predictive testing for Huntington’s, but it belongs in any genetics guide because it was for decades the single biggest reason Australians at risk of HD did not get tested.

Since 8 October 2026, Australian life insurers may not solicit or use adverse genetic test results in underwriting. The ban was legislated in April 2026 as an amendment to the Insurance Contracts Act, inserting section 33H, with a civil penalty of up to about $1.65 million. It covers death, total and permanent disability, trauma and income protection cover.

What it does not cover. Insurers can still consider existing diseases, signs or symptoms, and family medical history. So for somebody with an affected parent, the family history remains usable even though a predictive test result does not. And health insurance was never affected in Australia, because community rating means genetic information cannot be used there at all.

Older Australian material — including the 2017 Centre for Genetics Education booklet, which is otherwise excellent — states that genetic test results must be disclosed when applying for new life, trauma or income protection cover. That guidance is now out of date. If you are reading something about genetic testing and insurance in Australia that does not mention October 2026, check its date.