Paper: "Non-suicidal self-injury and its relation to suicide through acquired capability: investigating this causal
mechanism in a mainly late-diagnosed autistic sample"
This paper was published in 2022, in the journal Molecular Autism.
You can click HERE to download a PDF of the paper.
This study follows on from some previous research on the same topic. You may like to read the first and second studies on this topic first (they will open in new tabs). When you are ready, see below for a plain English summary and explanatory videos.
Why is this an important issue?
Many autistic people are affected by, or engage in, non-suicidal self-injury (NSSI). NSSI is when people deliberately hurt themselves to cause pain or injury to their body.
Our previous research suggests that autistic people engage in NSSI for many reasons. Some do not find it a problematic aspect of their lives, but others find it very distressing. NSSI can have negative consequences, such as when wounds or scars affect a person’s confidence or relationships. Most worryingly, though, something about engaging in NSSI seems to increase a person’s risk of suicide.
To explain the link between NSSI and suicide, psychologists put forward the concept of ‘acquiring capability’ for suicide. Typically, we are driven by evolutionary instincts to protect ourselves and avoid pain and danger to our lives. When a person ‘acquires capability’ for suicide, they are able to overcome these instincts if they wish to end their own life.
Psychologists suggest that NSSI builds capability for suicide because it often leads to people becoming habituated to pain, such that they are less frightened of pain in future and can stand to hurt themselves more severely. Hurting yourself also allows a person to ‘mentally rehearse’ possible ways of ending their life, such that they become used to these thoughts and death is also less frightening. These are ideas developed in the general population, but we don’t know if they are also true in autistic people.
You might ask: why is it useful to understand how NSSI increases suicide risk, if we can just accept that it does?
Essentially, it is about giving us the best chance to identify individuals at highest risk of suicide and prioritise helping them before it’s too late. It is possible that not all forms of NSSI are associated with suicide. If certain behaviours are especially dangerous and we understand what it is about them that makes them so, it may help us identify other risk factors which share something in common. Understanding what happens when someone engages in a certain behaviour, what unknown impacts it might have, can give us insights into a range of wider behaviours that we haven’t yet studied but which may have the same effect.
Here is a short (~6 minute) video summary. You can toggle your view by clicking the icon in the top right corner, and turn captions on and off by clicking the 'CC' button. See below for a fuller video explanation, if you'd like more details!
If you would like a more detailed explanation, please see the video below. You can still toggle your view by clicking the icon in the top right corner. Click 'CC' to turn captions on and off.
What was the purpose of this study, and what did the researchers do?
To better understand the relationship between NSSI and suicide, we asked 314 autistic and 312 non-autistic people to complete an online survey about their history with NSSI and suicidality. They also completed a questionnaire measuring the different aspects that mean someone has acquired capability for suicide: these include having high pain tolerance, being less frightened of pain and death, and being able to mentally rehearse suicide plans.
We aimed to look at relationships between NSSI and suicide attempts in autistic and non-autistic people. In particular, we wanted to explore if this relationship was explained by acquired capability for suicide - i.e., if people who engage in NSSI have higher rates of suicidality because they have acquired capability for suicide. You can imagine it as a two step pathway (see image below), where NSSI builds capability for suicide, and then that acquired capability for suicide is what increases their risk of attempting suicide.
In science, this is called mediation: if something is the mechanism which connects two things, we say that it mediates the relationship between those two things, such that they are indirectly linked through it.
As well as looking at the relationships between NSSI, acquired capability and lifetime suicide attempts, and whether these differed between autistic and non-autistic people, we looked at certain NSSI behaviours in detail. The three most common behaviours in both autistic and non-autistic people were severely scratching oneself, cutting oneself, and hitting oneself. We looked at whether the pain associated with NSSI was linked to how much people habituate to NSSI (i.e. becoming less disturbed by it and needing to do more to gain the same effect). We then looked at relationships between these methods, acquired capability, and suicide.
What were the results of the study?
The majority of our autistic sample (71%) had lifetime history with NSSI; so did 50% of our non-autistic comparison group. Both autistic and non-autistic people who self-harmed had a higher incidence of suicide attempts. The more people had engaged in NSSI, the more suicide attempts they tended to have.
This relationship was indeed partly explained by acquired capability for suicide. People with the highest rates of NSSI tended to have highest capability for suicide: that is, they were the least afraid of death, had the highest pain tolerance, and were most likely to mentally rehearse suicide plans. Having reduced fear of death and mentally rehearsing suicide plans, two aspects of acquired capability, were associated with having more numerous lifetime suicide attempts. In this way, NSSI was related to suicide attempts because it was associated with acquired capability. However, since acquired capability only partly bridged the relationship between NSSI and suicide, it is likely that NSSI is linked to suicide in additional ways. We know, for instance, that engaging in NSSI can worsen pre-existing mental illness; that NSSI can put a strain on relationships; and, over time, NSSI can result in people becoming more emotionally dysregulated. We did not measure them, but these and other factors could be the missing ‘stepping stones’ - i.e. the route through which NSSI increases suicide risk.
Were there any differences between autistic and non-autistic people? Interestingly, the relationship between NSSI and suicidality was weaker in autistic people. In science, we test the strength of relationships (or correlations) by asking how reliably one thing predicts the other: in this case, how often does NSSI occur in someone who has also attempted suicide? That the relationship between NSSI and suicidality was weaker in autistic people means that NSSI was less reliably linked to suicidality: autistic people were more likely to be suicidal whether or not they engaged in self-harm. In just the same way, autistic people tended to have greater capability for suicide whether or not they self-harmed. This is consistent with autistic people having higher suicide rates generally, and suggests that there are a wider array of factors that might be important for predicting suicide in autistic people.
We were not able to compare methods of self-harming between autistic and non-autistic people, and our results here were very tentative (see below for the problems with this analysis!). Compared with severely scratching oneself, hitting and cutting oneself were more strongly linked with suicide. The extent to which people habituated to NSSI was important for how linked these behaviours were to acquired capability and to suicide attempts. Pain experienced during NSSI was not associated with suicidality: that is, causing yourself greater pain did not predict a stronger link between that NSSI behaviour and acquired capability for suicide.
What are potential weaknesses in the study?
This study measured people’s thoughts, emotions and behaviours at one snapshot in time. We were working off a theory that suggests that NSSI causes people to acquire capability for suicide, which then results in increased risk of suicide attempts. However, because we tested participants at just one point in time, we cannot say whether participants actually began engaging in NSSI before or after the point where they had attempted suicide. Likewise, we cannot tell whether higher scores on acquired capability preceded and contributed to suicide attempts, or whether surviving a suicide attempt itself caused people to score higher on this poorly understood psychological concept.
We only looked at three forms of NSSI, and it is possible that the vague wording of the “scratching” item is the reason why this behaviour was seemingly less related to suicidality. We expected pain to be an important element of what might make some NSSI behaviours especially worrying. That we did not find this is likely because of weaknesses in the way we measured it - we asked about pain experienced during NSSI more generally (rather than pain associated with particular behaviours), and it is probably difficult to provide an accurate rating of pain you’re not experiencing in the present moment. In general, we were forced to use psychological questionnaires which weren’t designed with autistic people in mind, so these may not have worked totally as expected.
Our comparison between autistic and non-autistic people was limited because our non-autistic group were significantly younger (approx. 21 years old, while autistic participants were on average in their early 40s). This means we were not fully ‘comparing like with like’: other differences between the two groups, such as their age, might be the reason why autistic people, for instance, had lower fear of death.
Our autistic group was predominantly white, mainly female, and nearly all participants were diagnosed as adults. That means the findings might not be true for autistic people diagnosed as children; autistic people of colour; autistic cisgender men, as well as trans men, trans women, and people with other minority sex or gender identities. Autistic people with intellectual (learning) disabilities were also not represented.
How will these findings help autistic adults now or in the future?
Many, many autistic people will experience suicidal thoughts and feelings during their lifetime: a desperate, distressing, devastating experience. Suicide is one of the leading causes of death in the autistic community, with autistic people at much higher risk than the general, non-autistic population.
We know that autistic people also have markedly higher rates of NSSI, and so we wanted to understand whether these two things are linked: whether the high likelihood that autistic people will experience NSSI is associated with why they might have higher risk of dying by suicide.
Suicide is a complex phenomenon and there is no single cause or reason, or no single combination of reasons, which could explain and predict every situation. Nevertheless, our findings suggest that NSSI is associated with higher risk of suicidality, and offer a partial explanation for why engaging in NSSI could increase a person’s future risk of suicide. Specifically, this research suggests that NSSI affects a person’s ‘capability’ for suicide: i.e. how comfortable they are with thoughts of hurting themselves or dying. However, the research also shows that there are a broader range of factors involved in autistic suicidality, perhaps more so than in the general population, and that NSSI might impact suicide risk through a number of ways.
We hope that these findings will highlight to professionals the importance of supporting autistic people who want to reduce or prevent their self-harm. We also hope the findings will help professionals identify autistic people at highest risk of suicidality.
Thank you for reading!
If you found this interesting, you may like to read:
If you are struggling with feeling suicidal, I am deeply sorry. Please, please look at the resources page, just in case there is anything helpful for you