There has been a significant rise in the number of self-injury incidents reported in the UK’s mental health services over the last decade (Woodnutt et al., 2024). The way someone injures themselves can depend on where they are at the time. This is because access to things that can cause harm — like blades, glass, or other sharp or dangerous objects — affects how serious the injury might be. This idea is called access to means.
Limiting access to means has long been part of how governments try to prevent harm, for example by controlling who can have firearms, limiting access to strong medicines and toxic chemicals, or in hospital environments such as secure wards, removing personal items that could cause harm (Sarchiapone et al., 2011).
With the overall rise in self-injury across the population, mental health staff are seeing more patients who harm themselves, especially in secure wards. When harmful items are restricted, people may turn to other methods such as headbanging — repeatedly hitting their head against a hard surface like a wall or floor. Headbanging is one of the most common ways people harm themselves (Chester & Alexander, 2018), but there is still little research about why it happens, how often it occurs, or what situations lead to it.
This study undertook a review of headbanging and self-injury incidents across the last five years within a private sector secure mental health provider.
There has been a rise in self-harm incidents involving headbanging – particularly in secure wards where dangerous items are confiscated.
Methods
Doyle et al. (2025) carried out a study looking at how often headbanging and other self-injury incidents were reported in one of the UK’s largest independent mental health providers, which operates secure and specialist services for adults, adolescents and people with neurodevelopmental conditions. They included all data recorded over a five-year period. Their study included information from 421 patients whose incidents had been recorded according to hospital policy.
The researchers looked at a few key details about the people involved — their diagnosis, age, gender, and the type of ward or service they were in. This included places such as Child and Adolescent Mental Health wards, low secure wards, and other specialist environments.
Rather than counting the total number of headbanging incidents (because one person might have many, which can make the data hard to compare), the team focused on which patients had at least one headbanging incident during the study period. They then compared the characteristics of those patients to the characteristics of patients who had engaged in any other form of self-injurious behaviour to look for patterns; for example, whether certain age groups or diagnoses were linked to headbanging more often.
The study used the hospital’s internal incident reporting system, where staff record both the type and severity of each incident. Staff responding to an incident rated how serious it was using categories such as no harm, low harm, moderate harm, or serious incident. The more serious cases sometimes required the person to be transferred to a physical (general) hospital for treatment.
Results
Headbanging incident frequency
The results showed that most patients had fewer than 49 headbanging incidents over the five-year period, with the majority reporting between 0 and 9 incidents each. A small number of patients had between 50 and 99 incidents of headbanging, and an even smaller minority of patients had between 100 and 499 incidents of headbanging.
Most of the incidents of headbanging across the sample were rated as either ‘no harm’ or ‘low harm’. Only a small number of patients experienced moderate harm 10–49 times, while most had between 0 and 9 episodes of harm. Serious incidents only seemed to affect a very small number of patients. The authors do not report the exact figures, only the overall patterns in the form of a graph.
The authors also reported overall self-injury for the sample (which includes the headbanging data) and this showed similar trends across age, diagnosis and gender.
Characteristics of the patients in the sample
The authors reported that 64% of the headbanging incidents involved female patients (70% for self-injury) and that younger patients had more incidents of self-injury, while the number of incidents decreased with age. Only a very small proportion of 10-14 year old patients reported headbanging.
Most of the headbanging and self-injury incidents occurred in patients who had a diagnosis of Emotionally Unstable Personality Disorder, with much smaller amounts in other diagnoses such as Post-Traumatic Stress Disorder, Autism, Schizophrenia or a range of other mental health or neurodevelopmental disorders.
Finally, the authors looked at the type of service that the incidents came from. Most of these were from the Low Secure and Specialist Rehab wards, followed by Child and Adolescent Mental Health wards, Autism and Learning Disability Wards, Medium Secure Mental Health Wards and Neuropsychiatry wards. A very small minority of patients were reported to be from a community setting.
Those that engaged in headbanging as a form of self-injury tended to do so less than 10 times, and most incidents were rated as low harm by hospital staff.
Conclusions
This study presents a descriptive overview of headbanging and self-injury among a small, specialist group of patients in a private mental health service. The authors found that younger people, particularly women with a diagnosis of Emotionally Unstable Personality Disorder (EUPD), were most likely to engage in this form of self-harm. The highest rates were reported in Child and Adolescent Mental Health Services (CAMHS) and low secure wards.
While these findings provide a snapshot of headbanging incidents within one organisation, they sit within broader trends in self-injury. National surveys show that younger adults are more likely than older adults to report self-harm and to have conditions such as PTSD or ADHD (Adult Psychiatric Morbidity Survey, 2023/4), which can produce behaviours that overlap with or are sometimes interpreted as EUPD (Sarr et al., 2025). This suggests that the patients in this study reflect patterns already seen more widely.
Overall, the study adds a small, local view to an already well-established picture, that self-injury among younger women with EUPD remains a significant challenge across services.
This study of hospital records found that headbanging as a form of self-injury was more likely to occur in young females with a diagnosis of emotionally unstable personality disorder.
Strengths and limitations
One strength of this study is that it used routinely collected hospital data rather than relying on separate research interviews or questionnaires. Although these kinds of data can be affected by differences in how staff report incidents, they do reflect how incidents are recorded in real time and in line with hospital policy. The study also looked at records covering five years, which provided a reasonably large sample of incidents and allowed for some comparison over time.
However, there are several important limitations. This study only looks at basic statistics from a small group of patients, many of whom had repeated headbanging incidents. Because of this, it’s hard to know if the patterns they report apply to other patients or other hospitals.
The study mostly just describes who had incidents — their age, gender, and diagnosis — without trying to explain why headbanging happened or whether these factors really made a difference. It doesn’t use more advanced analysis that could show connections between patient characteristics, the environment, or staff practices.
The main findings — that headbanging was more common in younger, female patients with Emotionally Unstable Personality Disorder (EUPD) — aren’t new. Other research already shows similar patterns, so the study doesn’t really add new knowledge.
Finally, the study doesn’t give much context about how the ward environment or care practices might affect headbanging. For example, rules about what items patients can use, staff responses, or developmental factors could all influence behaviour, but the study doesn’t explore these. Without this context, it’s hard to interpret the results.
This is a small descriptive study about headbanging and self-harm – the study does not attempt to explore why headbanging occurred in particular populations.
Implications for practice
Future research and clinical practice would benefit from a more nuanced approach to understanding behaviours such as headbanging. Differentiating between self-harm, sensory regulation, and communicative behaviours could help avoid over-pathologising all instances of headbanging as self-injury. Recognising these distinctions is important for tailoring support strategies and for ensuring that interventions are appropriate to each individual’s needs and intent.
It would also be valuable to explore how environmental restrictions, emotional regulation, and developmental stage interact to shape patterns of self-injury within secure and inpatient settings. Considering these factors together could provide a more complete understanding of why certain behaviours occur, and how care environments might be adapted to reduce distress and promote safer coping strategies.
Future research should aim to differentiate between headbanging as self-harm, sensory regulation and communication in order to avoid over-pathologising all instances of headbanging as self-injury.
Statement of interests
have no conflicts of interest to declare.
Edited by
Laura Hemming.
Links
Primary paper
Isobel Doyle, Kristina Brenisin & Kieran C Breen (2025). The incidence of headbanging as a form of self-harm among inpatients within a secure mental health setting–the impact of age, gender and diagnosis. The Journal of Forensic Psychiatry & Psychology, 36(5), 697-710. DOI: https://doi.org/10.1080/14789949.2025.2545206
Other references
Adult Psychiatric Morbidity Survey: Survey of Mental Health and Wellbeing, England, 2023/4 – NHS England Digital
Chester, V., & Alexander, R. T. (2018). Head banging as a form of self-harm among inpatients within forensic mental health and intellectual disability services. Journal of Forensic Psychiatry & Psychology, 29(4), 557–573. https://doi.org/10.1080/ 14789949.2018.1425472
Sarchiapone, M. et al. (2011) ‘Controlling access to suicide means‘, International journal of environmental research and public health, 8(12), pp. 4550-4562.
Sarr, R., Spain, D., Quinton, A.M., Happé, F., Brewin, C.R., Radcliffe, J., Jowett, S., Miles, S., González, R.A., Albert, I. and Scholwin, A., (2025). Differential diagnosis of autism, attachment disorders, complex post‐traumatic stress disorder and emotionally unstable personality disorder: A Delphi study. British Journal of Psychology, 116(1), pp.1-33.
Woodnutt, S. et al. (2024) ‘Analysis of England’s incident and mental health nursing workforce data 2015–2022‘, Journal of Psychiatric and Mental Health Nursing, 31(5), pp. 716-728.
