ADHD and suicide attempts: the genetics adds to the red flags, it does not replace them
A large Danish study published this week looked at suicide attempts in five neurodevelopmental and psychiatric conditions. It asked whether genetic risk works the same way in each one. For ADHD, the clearest message is not about genes. It is about sex and substance use. The genetics adds to those.
The study
The team used iPSYCH, a Danish group of 98,069 people. Suicide attempts were taken from national hospital records, using ICD-10 codes (the international list of diagnoses). People with ADHD, autism, schizophrenia, bipolar disorder and depression were compared with people from the general population. A person could have more than one diagnosis.
The researchers used 21 polygenic scores. A polygenic score adds up thousands of small, common genetic differences linked to a trait. They split each diagnosis group into thirds by score. Then they worked out the absolute risk of a recorded suicide attempt in each third.
What they found
Recorded suicide attempts were 1% in the comparison group. In the diagnosis groups they ranged from 5% in autism to 31% in schizophrenia.
The genetic picture was broadly the same across conditions. People who had made an attempt had higher polygenic scores for ADHD, depression and substance use, whatever their diagnosis.
ADHD: sex and substance use come first
In ADHD, females in the top third of the score had a 36% risk of a recorded suicide attempt. Males in the top third had 22%.
Substance use raised risk further. In people with ADHD and a substance use disorder, risk reached 48% in the top third of the score.
The scores did not explain the sex difference. Females and males who had made an attempt carried much the same genetic load. The genetics shifts risk within each group. It does not explain the gap between them.
Autistic young people
Autism had the lowest rate of recorded attempts of the five conditions, at 5%. That is still five times the rate in the comparison group.
The pattern in autism matched the other groups. Autistic people who had made an attempt had higher scores for ADHD, depression and substance use. So in an autistic young person, ADHD, low mood and alcohol or drug use need the same attention they get in ADHD.
What the scores can and cannot do
Polygenic scores are not a clinical test. They describe risk across groups. They cannot tell you whether a particular young person will make an attempt.
Hospital records miss some attempts. The authors suggest scores may help risk assessment in future. That is not where practice is now.
In clinic, the questions that matter are still the plain ones: mood, sleep, alcohol, drugs, and what has changed recently. I would not order a polygenic score for this.
Take-home
- In ADHD, being female and using alcohol or drugs are the clinical red flags.
- Ask young people with ADHD about alcohol and drugs directly, and keep asking as they get older.
- In autistic young people, look for ADHD, depression and substance use as well.
- Genetics adds to clinical risk. It does not replace clinical assessment.
- In every group, most people did not make a suicide attempt. Risk can be spotted and acted on.
Reference
- Starnawska A, Nielsen TT, Grove J, Cordero-Solorzano J, Bybjerg-Grauholm J, Psychiatric Genomics Consortium Autism Spectrum Disorder Working Group, Corfield EC, Anney RJL, Docherty AR, Børglum AD, Demontis D. Genetic architecture and risk of suicide attempt across neurodevelopmental and psychiatric conditions. Biological Psychiatry. Published online 13 October 2026. https://doi.org/10.1016/j.biopsych.2026.09.028 (open access).
If this article raises difficult feelings for you or someone you care for, support is available. In the UK, call Samaritans free on 116 123, day or night, or text SHOUT to 85258. In England, call NHS 111 and choose option 2 for urgent mental health help. If a life is at risk, call 999.