There’s no getting around it, self-harm and related thoughts and behaviours are a big deal, with one in four adolescents reporting ever having harmed themselves on purpose (Patalay & Fitzsimons, 2021). Not only causing physical injury, self-harming thoughts and behaviours can cause high levels of distress and impact on the young person and those around them.
One of the worst possible outcomes is suicide, and although suicide is thankfully rare (around 9.09 per 100,000 15-24 year olds in the UK) it is very hard to predict who is most at risk. Although most young people who self-injure do not go on to die by suicide, self-harming behaviours are one of the strongest predictors of suicide risk (Hawton et al., 2007). Naturally it makes sense to understand the mechanisms that increase these risks, so that we can prevent both self-harm and suicide-related outcomes.
A lot of research to date has focused on psychological risks, most notably depression (Knipe et al., 2022; Wasserman et al., 2021). However, emerging research is beginning to show that biology may play an important role too, although not in a simple ‘gene X = outcome Y’ manner, more like “genetic variation in genes 1 to 2000=50% of variance in outcome Y explained” (Lim et al., 2022).
Self-harm is common among adolescents and strongly linked to suicide risk, prompting growing interest in both psychological and genetic factors that could inform prevention strategies.
Methods
A new study published in the British Journal of Psychiatry (Wen et al., 2025) uses data from nearly 5,000 9-10 year old children who had no history of self-harm, and followed them up for four years to see who started to have thoughts of self-harm, actual self-harm and suicide attempt (I refer to all these as “self-harm” from now on).
The researchers looked at whether genetic risk or brain structure and function predicted self-harm risk and whether different genetic susceptibility – measured by something called a polygenic risk score (which is derived from adding up risk measured across many genetic variants that have been determined through separate research studies), was associated with different trajectories of the self-harm outcomes. They also looked at whether the link between genetics and self-harm was explained by brain structure or connection (from MRI scans). They used what we would consider ‘best practice’ research methods.
Results
Over the four years of the study, more and more of the sample reported self-harm, with 29.3% reporting this by the final follow-up. Females from disadvantaged households were the most at risk. The authors found that adolescents with higher polygenic risk scores had increased odds of reporting self-harm. The only exception was the risk score for ‘non-suicidal self-injury’. The genetic risk scores for ever having thought about or self-harmed and ever attempted suicide were all linked to higher odds of young people’s self-harm over four years. This effect looked like it got stronger over time (the odds ratios increased), but there was still a big overlap in the confidence intervals, so we can’t say this strengthening over time was true with certainty. Most odds ratios were around 1.2.
The researchers also looked at whether the trajectory of self-harm over time was different for those with low versus high genetic risk. These findings are hard to interpret as the data only cover 5 years, but in general the ‘high genetic risk’ group had a more prominent increase in risk.
In terms of the brain, the researchers found big differences in the amount of grey matter (areas of the brain that have a lot of cell bodies) in one area that is involved in communication across the brain (the left ventral posterior cingulate cortex for those neuroanatomy buffs out there), with an odds ratio of 1.17 (95% CI 1.07 to 1.28). There were also indications of differences in grey matter in six other regions. Some differences were found in connections between specific brain areas and networks relating to attention, as well as networks that are active when we are at rest (default mode network).
The research team found some evidence that differences in brain structure and connection explaining a small amount of the link between genetics and self-harm, with generalised mental health also playing a role. Interestingly the mediating pathways that were statistically significant did not involve the same brain regions/connections that were implicated in the strongest direct associations mentioned above.
Adolescents with higher genetic risk scores showed increased odds of self-harm over time, especially among disadvantaged females, with brain structure and connectivity differences offering partial explanations alongside general mental health.
Conclusions
Overall, the findings mean that there is evidence that genetic risk for self-harm and suicide attempt does link to an increased risk of young people thinking or acting on self-harm in adolescence in this cohort. However, genetic risk for non-suicidal self-injury does not appear to play a role.
In addition to this, differences in the volume of grey matter in several brain regions (one in particular) is apparent in those who do versus do not self-harm, as well as some of the connecting networks across areas of the brain. Interestingly, this includes the ’default mode network’ which is active when we are at rest. This suggests that there is an important biological component to adolescents’ risk of self-harm, although not of the magnitude where genetic testing for predisposition would be of use at the moment.
Genetic risk for self-harm and suicide is linked to adolescent self-harming behaviours, with brain structure differences pointing to a biological component. However, genetic testing isn’t yet at a stage where it can guide prevention.
Strengths and limitations
Cohort studies are very powerful for studying risk factors because they sample a good variety of the general population (rather than, for example, only those who go to hospital following self-harm, which we know is a very small proportion of those who actually self-harm). They also help us to make inferences about cause and effect, as they follow people over time and you can account for what happens first, and the impacts that follow. One of the downsides is that people drop-out over time, and of the ~12,000 who did the baseline measure, only ~4,000 were still answering the questionnaires four years later. However, this is still a good-sized sample.
Regarding the genetic risk, the one finding that was not statistically significant was the risk score for ‘non-suicidal self-injury’. This doesn’t surprise me as this is still a controversial conceptualisation of self-harm that encompasses those who self-harm but report no suicidal intent. Part of the reason this is controversial is that many people who self-harm do so repeatedly, and intent may vary from one episode to another. Therefore, creating a group who ‘only’ self-harm without any suicidal intent is a challenge in itself.
The authors make a claim that their data support the 10–15-year-old age window as a critical period for self-harm aetiology. However, I struggle to see how they can know this for sure, as their entire sample was this age and they excluded anyone with a history of self-harm before the baseline assessment. The sample overall was large, but there were still only several hundred individuals reporting self-harm by the final follow-up. Therefore, in order to make generalised conclusions, I’d like to see this finding replicated in other populations.
One other little-talked-about limitation of studies using genetic data is that you often have to restrict your sample to those with similar ancestry; in this case European. This means that we do not know if these findings apply universally to adolescents or if there are specific risks for different subgroups of the population.
Cohort studies offer strong population-level insights, but limitations of this study include participant drop-out, ancestry restrictions, conceptual challenges around non-suicidal self-injury, and uncertainty in defining critical developmental windows.
Implications for practice
Increasingly, research is uncovering potential biological causes or mechanisms linked to self-harm, although it seems clear that psychological factors still play an important role. As so many young people self-harm, we clearly need to understand as much as possible about the causes so that we can work to prevent its onset and continuation. This is important as some still dismiss self-harm as being “attention seeking” or attribute stigma to those who self-harm, who are already likely feeling isolation, distress and shame.
However, studies such as this one have quite a long way to go before there are clear implications for practice. As is often the case for epidemiological research, this study represents a piece of a much larger puzzle. It has, however, helped to narrow down brain regions that might be implicated and could be further studied. For example using case-control methods to explore activation of these same areas of the brain in those who self-harm versus those who do not. The genetic findings also need replicating and validating in larger and more varied samples, not just in those of European descent.
One important implication is to do with prevention of the onset of self-harm, not just treating the behaviour ‘after the fact’. There is potential for this study, in combination with others, to further identify genetic risk pathways or even (in the distant future) to be able to screen and identify a biological predisposition, so that prevention work can be targeted to those who are at highest risk.
Biological mechanisms linked to self-harm are emerging, but practical applications remain distant. Current findings highlight the need for stigma reduction, further brain and genetic research, and a stronger focus on early prevention.
Statement of interests
Abby Russell does not have any conflicts of interests relevant to this blog post.
Links
Primary paper
Wen, X., Sun, Y., et al (2025). Genetic and neurobiological mechanisms underlying transition in self-injury thoughts and behaviours during adolescence. The British Journal of Psychiatry, 1–13. https://doi.org/10.1192/BJP.2025.10300
Other references
Hawton K., & Harriss L., (2007). Deliberate self-harm in young people: characteristics and subsequent mortality in a 20-year cohort of patients presenting to hospital. Journal of Clinical Psychiatry 68(10) 1574 https://www.psychiatrist.com/read-pdf/4097/
Knipe D., Padmanathan P., et al (2022). Suicide and self-harm. The Lancet, 399(10338), 1903–1916. https://doi.org/10.1016/S0140-6736(22)00173-8
Lim, K. X., Krebs, G., et al (2022). Investigating the genetic and environmental aetiologies of non-suicidal and suicidal self-harm: a twin study. Psychological Medicine, 52(15), 3391–3401. https://doi.org/10.1017/S0033291721000040
Patalay, P., & Fitzsimons, E. (2021). Psychological distress, self-harm and attempted suicide in UK 17-year olds: prevalence and sociodemographic inequalities. The British Journal of Psychiatry, 219(2), 437–439.
Royal College of Paediatrics and Child Health (2020) State of Child Health. London: RCPCH. [Available at: stateofchildhealth.rcpch.ac.uk]
Wasserman, D., Carli, V., et al (2021). Suicide prevention in childhood and adolescence: a narrative review of current knowledge on risk and protective factors and effectiveness of interventions. Asia‐Pacific Psychiatry, 13(3). https://doi.org/10.1111/APPY.12452
