Mental health

Therapy reduces risk of suicide or self-harm

“Talk therapy sessions can help reduce the risk of suicide among high-risk groups,” BBC News reports.

The headline is prompted by a large Danish study that took place over a 20-year period.

Researchers matched those who had received different psychosocial (“talking therapy”) interventions after a self-harm attempt with those who had not received a psychosocial intervention, and then compared relevant outcomes.

People who received psychological interventions had reduced risk of further self-harm, but not suicide, within the first year. Looking at longer-term follow-up, psychological interventions were associated with reduced risk of both self-harm and suicide.

However, it may be difficult to isolate the direct effect of the psychological intervention. People who had received psychological interventions were recruited from treatment clinics that required them not to be in need of psychiatric admission.

Meanwhile, those who did not receive psychological treatment were reported to include people who needed psychiatric admission, or chose not to receive suicide prevention treatment. These factors could mean that this comparison group were at increased risk of subsequent harm and death to begin with.

Also, the situation in the UK might be slightly different to Denmark. Despite this, any research that could help prevent suicides is always valuable.

Where did the story come from?

The study was carried out by researchers from the University of Copenhagen in Denmark and the Johns Hopkins Bloomberg School of Public Health in the US, in addition to other research institutions in Denmark and Norway. Funding was provided by the Danish Health Insurance Foundation; the Research Council of Psychiatry, Region of Southern Denmark; the Research Council of Psychiatry, Capital Region of Denmark; and the Strategic Research Grant from Health Sciences, Capital Region of Denmark.

The study was published in the peer-reviewed medical journal The Lancet Psychiatry.

BBC News was generally representative of the research’s findings, but inaccurately described participants as having “attempted suicide”. The research included participants who had self-harmed. Not all instances of self-harm are suicide attempts, so it is a mistake to conflate the two terms. For some people, certain types of self-harming, such as cutting, are a way of coping with overwhelming emotional distress, rather than an attempt to end their life.

It was not clear from the study what proportion of the self-harming events were attempted suicide.

What kind of research was this?

This was a cohort study comparing people who did and did not receive a psychosocial (talking) therapy after deliberate self-harm, and examined the outcomes of further self-harm, suicide or death from other causes. 

The researchers say that self-harm is a strong predictor of suicide. Research indicates that within the first year after self-harming, about 16% of people self-harm again; 0.5 to 1.8% die by suicide; and 2.3% die from another cause. However, evidence for the effectiveness of psychological interventions following self-harm is said to be missing, and this study aimed to investigate this.

What did the research involve?

This study compared people in Denmark who received a psychological intervention following a first episode of self-harm with those who received standard care, over the 18-year period between January 1992 and December 2010. They calculated the risk of repeated self-harm, suicide and dying of any cause after the first instance of self-harm, and compared the risks between the two groups for differences that might be due to the psychological intervention. 

The people who received psychological interventions were identified from one of seven suicide prevention clinics in Denmark. These clinics are said to receive people who are thought to be at risk of suicide, but not in need of psychiatric admission or other outpatient programmes. For the purposes of this study, participation was considered to be attendance for at least one psychological treatment session that was focused on suicide prevention. The seven different clinics used various types of therapy, including cognitive, problem-solving, crisis, dialectical behaviour, integrated care, psychodynamic, systemic, psychoanalytic approaches and support from social workers.

The controls who did not receive a psychological intervention were people who had presented to hospital with an episode of self-harm during the study period, but who did not receive any psychological intervention. They could receive any form of standard care, including admission to a psychiatric hospital, referral to outpatient treatment or a general practitioner, or discharge without referral.

The reasons why these people did not receive a psychological intervention were variable, including:

  • living in an area remote from services
  • being referred for other treatment (including hospital admission)
  • not wanting to be referred for suicide prevention treatment

All people were linked via their Danish ID numbers to the Danish Civil Register, National Registry of Patients, Psychiatric Central Registry and Registry of Causes of Death. Follow-up was to the end of 2011, giving a follow-up period for the people in the study of 1 to 20 years.

The main outcomes examined were self-harm, death by suicide, and death by any cause. People who did and did not receive psychological interventions were matched for various potentially confounding factors, including:

  • study period (1992 to 2000 or 2001 to 2011)
  • age
  • gender
  • educational level
  • socioeconomic status
  • previous episodes of self-harm
  • specific psychiatric diagnoses

What were the basic results?

The study included a total of 5,678 people in the psychological intervention group and 17,034 matched people who had not received a psychological intervention after self-harm. Around two-thirds were women and most were in the 15 to 49 age bracket. Around 10% had a previous episode of self-harm.

During the first year of follow-up, 6.7% of people receiving a psychological intervention had a repeated self-harm attempt, compared with 9.0% of the no psychological intervention group. Psychosocial therapy was associated with a 27% reduced risk of self-harm within one year (odds ratio (OR) 0.73, 95% confidence interval (CI) 0.65 to 0.82). The absolute risk reduction (ARR), measuring how much the risk of self-harm is reduced in those who received the psychosocial therapy, was 2.3% (95% CI 1.5 to 3.1%). The number needed to treat (NNT) was 44 (95% CI 33 to 67), indicating that 44 people would need to receive psychosocial therapy after a self-harm attempt to prevent one person self-harming within one year.

There was no significant difference between groups in rates of suicide within one year, but overall mortality rates within one year were slightly lower in the psychological intervention group (1,122 compared with 1,824 per 10,000), which also meant a significant reduction in overall mortality rate (OR 0.62, 95% CI 0.47 to 0.82). When considering the longer term effects over the full 20 years of follow-up, psychological intervention was associated with a 16% decreased risk of repeated self-harm (OR 0.84, 95% CI 0.77 to 0.91), with an ARR of 2.6% (95% CI 1.5to 3.7) and NNT of 39 people (95% CI 27 to 69).

When looking at overall follow-up, psychological therapy was also associated with a 25% reduced risk of death from suicide (OR 0.75, 0.60 to 0.94), with an ARR of 0.5% (95% CI 0.1 to 0.9) and a NNT of 188 people to prevent one suicide (95% CI 108 to 725). It was also associated with significant reduction of death from any cause (OR 0.69, ARR 2.7%, NNT 37).

The results altogether suggested that during the 20 years of follow-up, 145 self-harm episodes and 153 deaths were prevented by psychological interventions, with 30 of these deaths from suicide.

How did the researchers interpret the results?

The researchers conclude that their findings, “show a lower risk of repeated deliberate self-harm and general mortality in recipients of psychosocial therapy after short-term and long-term follow-up, and a protective effect for suicide after long-term follow-up, which favour the use of psychosocial therapy interventions after deliberate self-harm”.


The researchers report that this is the largest follow-up study of psychosocial interventions offered after deliberate self-harm attempts. Compared to standard care, it found that psychosocial interventions were associated with a reduced risk of repeated self-harm and death from any cause within the first year of follow-up. In the longer term, psychosocial interventions were associated with reduced risks of self-harm, death from any cause and suicide, specifically.

The study benefits from its large sample size, long duration of follow-up and reliable methods of identifying participants and their outcomes. There are, however, some points to be considered when interpreting the findings.

Possible selection bias

The reasons that people did not receive a psychological treatment could have put them at higher risk of subsequent harm to start with, potentially explaining all or some of the risk difference between the two groups. Though the people who did and did not receive psychological treatments were matched for various factors, this may not have been comprehensive, and some selection bias may still be present. For example, all the people who were receiving psychological treatments had been referred to suicide prevention clinics because they were not considered to be in need of psychiatric admission or other outpatient treatment following their self-harm attempt. Meanwhile, those who did not receive psychological treatment were reported to include people who needed psychiatric admission, or chose not to receive suicide prevention treatment after their self-harm attempt.

This makes it difficult to isolate the effect of the psychological intervention compared with selection biases and other confounding factors. It could be that the reduced risk seen in the psychological intervention group is not solely a result of the intervention, but that there were other risk factors among the non-treated group that were increasing their risk of further self-harm/suicide attempts and so confounding the association.

However, some degree of selection bias is inevitable in this type of study. The only way to remove it completely would be to randomise people to treatment or no treatment, which could never be done for ethical reasons.

Uncertainty about most effective intervention

It is also difficult to conclude many treatment implications from this study in terms of what would be the best type of psychological intervention to use after a self-harm attempt (a wide variety of interventions were used in this study), whether the optimal type differs according to the individual (e.g. according to mental health diagnosis[es]), and what would be the optimal treatment duration.

Results may not be applicable to the UK

The results also apply to Denmark, which may differ from other countries – for example, in terms of healthcare and mental health services, and population health, psychosocial and environmental influences. This may mean that the results are less applicable to this country.

People in the UK who present to health services following self-harm or a suicide attempt receive assessment by specialist mental health professionals, followed by referral, hospital admission or discharge, and follow-up care and treatment as appropriate to their individual situation.

Getting help

If you are reading this because you are having suicidal thoughts, try to ask someone for help. It may be difficult at this time, but it's important to know you are not beyond help and you are not alone.

Speak to a person you trust (such as a friend or family member), make an urgent appointment with your GP or contact your local A&E department. The Samaritans (08457 90 90 90) also operates a 24-hour service available every day of the year.

Read more about getting help for suicidal or self-harming thoughts, as well as spotting possible warning signs in family members and friends.

NHS Attribution