Part 1 · The disease
Genes: C9orf72, SOD1 and more
A change in a single gene explains some ALS, in families and in people with no family history at all. Pick a gene and press Play to see what goes wrong inside the cell. Then follow a family through the years, compare the genes, and read how the 2023 US consensus describes each kind of genetic result.
Schematic, not to scale. The repeat count on the dial is one example allele; the coloured ranges come from the studies cited in the notes. Gene, exon and protein sizes are drawn for clarity.
How much ALS runs in families?
A meta-analysis of 33 studies put familial ALS at 4.6% (95% CI 3.9–5.5%) of all ALS, and 5.1% (4.1–6.1%) in prospective population-based registries. The authors concluded it is "not 10% as is often stated". Rates were lower in southern Europe. 86
The 2022 Lancet Seminar counts ALS or related syndromes such as FTD among relatives, and says familial ALS may be underreported because of low-penetrance variants, small families and incomplete histories, perhaps "closer to 20% of cases". 180
Controversy Much of the gap is definition: whether relatives with FTD count, and how many generations are asked about. 86,180 The Seminar proposes replacing "familial versus sporadic" with genetically confirmed versus not genetically confirmed ALS: about 70% of familial cases, and about 15% of people with sporadic ALS, carry a pathogenic variant in a known gene. 180 In a UK clinic, 93% of people with a pathogenic or likely pathogenic variant had no family history of ALS. 466 Twin data put the heritability of sporadic ALS at 0.61 (95% CI 0.38–0.78), and an Irish parent-offspring study at 52.3%. 8,459
One copy is enough, but it is not a certainty
Most ALS variants are dominant: one changed copy, from one parent, can cause disease. Recessive and X-linked forms exist but are rare, and a person with a genetic form may have no recognised family history. 453 The C9orf72 expansion shows why: a carrier parent may never become ill, because penetrance depends on age and is incomplete. 213 Move the age slider, or press Play to let the family age, and switch between the two kinds of estimate.
An example family, not a real one. The chart shows how published studies estimate penetrance in groups of carriers; it is not anyone's personal risk, which is a matter for genetic counselling.
The genes side by side
By 2022 more than 40 genes had been reported in ALS. They differ in how common they are, how they are inherited (mostly dominant, rarely recessive) and how often carriers become ill; some, such as ATXN2, raise risk rather than cause disease. 180 In a meta-analysis of 111 studies, variants in the four main genes were found in 47.7% of familial and 5.2% of sporadic ALS, with a different pattern in Asia. 554
Share of familial and of sporadic ALS carrying a pathogenic variant in each gene, by region, in the 2017 meta-analysis. Bars share one scale. 554
| Gene | Inheritance | Share of ALS | Typical features | FTD link | ClinGen ALS panel |
|---|---|---|---|---|---|
| C9orf72 (GGGGCC repeat) | Dominant, age-dependent and incomplete penetrance | Europe: 33.7% familial, 5.1% sporadic; Asia: 2.3% and 0.3% 554 | Earlier onset, cognitive and behavioural change, shorter survival 87 | Strong: the commonest genetic cause of familial FTD 313 | Definitive 116 |
| SOD1 | Dominant; D90A (p.D91A) also recessive (homozygous) in Scandinavia 19 | Europe: 14.8% familial, 1.2% sporadic; Asia: 30.0% and 1.5% 554 | Limb onset, younger onset, survival varies by variant 410 | Cognitive impairment "not a prominent feature" 449 | Definitive (dominant) 116 |
| TARDBP | Dominant | Europe: 4.2% familial, 0.8% sporadic 554 | Variants in the C-terminal region 449 | Yes, ALS and FTD families 449 | Definitive 116 |
| FUS | Dominant; often de novo in young onset 238 | Europe: 2.8% familial, 0.3% sporadic; Asia: 6.4% and 0.9% 554 | Juvenile or young onset; rapid with P525L 356 | Not reported in the sources read | Definitive 116 |
| TBK1 | Dominant (loss of function, haploinsufficiency) 188 | About 1% of familial and 1% of sporadic 101 | Heterogeneous; extrapyramidal, ataxic and psychiatric features reported 101 | Yes 188 | Definitive 116 |
| NEK1 | Dominant susceptibility (risk variants) | Nearly 3% of ALS 259; up to 2% 101 | Typical ALS without dementia 101 | No, as reported 101 | Definitive, as "susceptibility to" 116 |
| OPTN | Recessive and heterozygous forms 323 | Rare in Europeans 449 | Also a glaucoma gene 323 | Not reported in the sources read | Definitive, semidominant 116 |
| VCP | Dominant | 1–2% of familial 251 | Part of IBMPFD: myopathy, Paget disease of bone, FTD 251 | Yes (IBMPFD) 251 | Definitive, curated for IBMPFD 116 |
| UBQLN2 | X-linked dominant 139 | Not a common cause 449 | Juvenile and adult onset 139 | Yes, ALS with dementia 139 | Definitive, X-linked 116 |
| KIF5A | Dominant | Rare (frequency not given in the abstract read) | Longer survival with loss-of-function variants 402 | Not reported in the sources read | Definitive, as "susceptibility to" 116 |
| ANXA11 | Dominant | 13 of 751 familial exomes 475 | Founder variant p.D40G 475 | Not reported in the abstract read | Definitive 116 |
| PFN1 | Dominant | 7 of 274 familial cases 544 | Aggregates containing TDP-43 in cells 544 | Not reported in the abstract read | Definitive 116 |
| CHCHD10 | Dominant | Rare 101 | Mitochondrial myopathy, cerebellar ataxia 39 | Yes, FTD-ALS 39 | Moderate 116 |
| ATXN2 (intermediate CAG repeat) | Risk factor, not a cause 180 | 4.7% of ALS had 27–33 glutamines (odds ratio 2.8) 157 | Long expansions cause spinocerebellar ataxia type 2 instead 157 | Not reported in the abstract read | Curated only for SCA2 (Definitive) 116 |
Shaded rows: the four genes every minimum test panel should include under the 2023 US consensus. 453 ClinGen ratings from the gene-validity file dated 2026-10-20; ratings are revised as evidence accumulates. 116
Reading a genetic result
The 2023 evidence-based consensus guidelines on ALS genetic testing and counselling came from a US expert panel, not an international one. It drew on a rapid review of 263 articles, GRADE and EGAPP evidence grading and a three-round modified Delphi process, and made 35 recommendations: 7 on clinical testing, 19 on counselling and 9 on laboratory methods. 453 Pick a result category to see how the panel describes it.
This explains published wording for each category. It does not interpret anyone's report; a genetic counsellor or clinician does that.
Offer testing to everyone with ALS
"All persons with ALS should be offered genetic testing" (GRADE A, strong). Before this, only 10–50% of clinicians offered testing in apparently sporadic ALS, US care guidelines did not address it, and European guidelines limited it to familial ALS or the SOD1 D90A phenotype. 453
Which genes
- At minimum C9orf72, SOD1, FUS and TARDBP. A first-round statement making C9orf72 the first test only for people of European descent failed; the revised one, covering everyone with ALS, reached 100% agreement.
- Further genes: those ClinGen rates as definitively associated with ALS, with "Limited", "Refuted" or uncurated genes clearly marked on the report.
- Simultaneous testing (panel, exome or genome) rather than one gene at a time.
- When the FDA approves a therapy aimed at a gene, everyone with ALS should be offered testing for that gene. 453
Counselling comes first
Counselling should be offered to everyone with ALS, whatever the family history, before the test. It includes at least a three-generation family tree recording ALS, PLS, PMA, FTD, other dementias, movement disorders and psychiatric illness, and covers inheritance, penetrance, risk to relatives, and psychological, economic and privacy effects, including discrimination. Ideally a board-certified genetic counsellor gives it; neurologists and nurse practitioners may when none is available. 453
Presymptomatic testing
The consensus refers to separately published recommendations from the Pre-fALS study (8 years; 317 counselling sessions for 161 first-degree relatives). They set a minimum age of 18 while no preventive therapy exists, require at least 2 counselling sessions, and rest on voluntary informed consent, psychosocial readiness, and counselling before the decision, before the test and after it. 48,453 Of 11 people who sought counselling before deciding, "over half" chose the study arm in which results are not disclosed; that was a research setting, not routine care. 48
Reproductive options
At-risk people may "pursue preimplantation genetic testing" (PGT) to avoid passing the variant on, and prenatal testing and PGT are possible once a family's C9orf72 expansion is known. 48,213 No published series on how often families with ALS use PGT was found for this guide.
GINA and its gaps
As summarised in 2016, the Genetic Information Nondiscrimination Act (GINA) bars health insurers and many employers from using genetic information, but no federal law covers life, disability or long-term-care insurance. 48 Laws differ between countries.
Israeli populations
C9orf72 in Ashkenazi and North African Jews
In Ashkenazi Jews the expansion was found in 80% of familial and 11% of sporadic ALS: 12.9% of all Ashkenazi ALS against 0.3% of controls (odds ratio 44.3). In North African Jews it was 10% of familial and 9% of sporadic ALS: 9.1% against 1% (odds ratio 9.9). All patients shared one risk haplotype. 209
Which test first?
In an Israeli neurogenetics clinic (140 people with ALS), the C9orf72 expansion was found in 21% of Ashkenazi Jewish patients and 5.7% of others. Gene panels found disease-causing variants in 14% of non-Ashkenazi and 5.7% of Ashkenazi patients, and in 12% of people with onset before 50, against 2.9% for C9orf72. 41
The authors propose C9orf72 first for Ashkenazi patients and a panel first for others and for early onset. This is a single-centre proposal, not a national guideline; the US consensus asks for the four-gene minimum in everyone. 41,453
Onset and duration by origin
Among 374 Israelis with sporadic ALS, people of North African Jewish origin were younger at onset and had a shorter illness than other groups. The authors suggest a genetic modifier but did not identify one. 148 No SOD1, FUS or TARDBP founder variant specific to Israeli or Jewish populations was found in the sources read for this guide.
Beyond one gene, one family
More than one variant
In a Dutch series, five familial ALS families carried several variants, more often than chance would predict. 505 In a UK clinic, 13% of people carried more than one variant, and those with two had earlier onset (median 56 vs 60 years). 466 A 2014 review argues that two "mutations" in someone from an outbred population more often mean one is not pathogenic. 449
New variants in the child
In 173 parent-child trios, the rate of de novo variants in ALS was no higher than in the general population. 509 Yet de novo FUS variants are the commonest finding in early-onset sporadic ALS: 6 of 14 people with onset before 35 carried one, and in all 5 families tested it was new. 238
Genes that change the odds
In UK Biobank, C9orf72 carriers homozygous for the UNC13A rs12608932 C allele had a higher risk (hazard ratio 1.81). 193 A cross-ancestry study found 15 risk loci and quoted a lifetime ALS risk of 1 in 350. 513