Can Sociopaths Be Fixed?

Two Cochrane reviews, both published in 2020, gathered the randomised trials of psychological and drug treatments. Cochrane is an independent, non-profit network that reviews health research. The psychological review identified 19 randomised trials covering 18 different interventions. Usable outcome data came from 10 of those trials, with 605 adults. The review’s conclusion was that there is not enough good quality evidence to recommend or reject any psychological treatment for people diagnosed with antisocial personality disorder (ASPD). The drug review covered 11 trials and 416 participants and concluded that the evidence is insufficient to draw any conclusion.

Those two conclusions are the state of the field. They mean the studies that would answer the question have not been done to an adequate standard. They are not the same as “nothing works.” In the psychological review, every comparison with usable data rested on a single trial and involved between 14 and 142 people. The drug review calls its evidence “single, unreplicated studies of mostly older medications”. It judged the risk of bias from dropouts high in 7 of its 11 trials. None of its trials set out to recruit people because they had ASPD, and many of the people in them came mainly with drug or alcohol problems. Of the 19 psychological trials, only four studied people with ASPD alone; the rest included them as a subgroup.

The outcomes measured are also narrower than the question. The psychological review’s main outcomes were aggression, reconviction, general and social functioning, and adverse events. Only four of the review’s 19 trials measured reconviction, and none of its outcomes was a change in the diagnosis itself. The review’s authors found that no treatment reported compelling evidence of change in antisocial behaviour. A trial can report an improvement without anything having changed about the underlying personality pattern.

The largest trial, Mentalization for Offending Adult Males (MOAM), came after both reviews. Published in The Lancet Psychiatry in 2025, it randomised 313 men on probation in England and Wales. The men had antisocial personality disorder (ASPD) and scored at least 15 on an aggression scale. Half had a year of mentalisation-based treatment added to probation, and half had probation alone. Mentalisation-based treatment aims to help people reflect on what they and others are feeling, and why. At 12 months, mean aggression scores were 90 among the treated men against 186 among the men on probation alone. The adjusted difference was −73.5, with a confidence interval, the range in which the true difference very probably lies, from −113.7 to −33.2. The authors describe the effect size of 0.74 as a medium-to-large effect.

The trial went on measuring after the year of treatment ended. The Mental Elf, a blog that summarises mental health research, has reviewed the trial. Its review reports that by 24 months the gap in aggression scores had narrowed to 25.4 points and was no longer statistically significant. It also reports that on a diagnostic interview for ASPD both groups scored lower a year in, the treated men more so. Over three years, it reports, the two groups’ offending was about the same, though offending was lower among the treated men in the third year alone. That year lay beyond the 24 months of offending records the trial’s published protocol set out to collect. MOAM is one trial, run at 13 sites with assessors who did not know which group each man was in. University College London’s press release on the study says its design made it hard to separate the therapy from the extra attention the men received. It also says the treated men committed 46% fewer offences than the men on probation alone over a three-year follow-up. The Mental Elf’s review gives a difference of that size for the third year alone. The trial’s authors say future research should test whether the gains last and apply more widely.

Everything else is weaker. Cognitive behavioural therapy added to usual care was tested against usual care alone in 52 men with ASPD and recent violence. It showed no difference in physical aggression, with an odds ratio of 0.92, where 1 would mean no difference. Its confidence interval ran from 0.28 to 3.07. Six sessions of impulsive lifestyle counselling and a post-release risk management programme also showed no difference from usual care. A drink-driving programme added to 28 days of incarceration showed none against incarceration alone.

The Cochrane review of psychological treatments judged three results to favour treatment, from contingency management, schema therapy and dialectical behaviour therapy. Schema therapy helps people replace long-standing, self-defeating patterns of thinking, feeling and behaving, and dialectical behaviour therapy trains skills such as tolerating distress. Contingency management gives people rewards for meeting a target such as a clean drug test. Its result was a small improvement in a social functioning subscale in one trial of 83 people, rated low certainty. The measure was drawn from an addiction severity index rather than from antisocial behaviour.

Schema therapy showed no difference in reconviction that its trial could detect. But in the same trial of 30 patients in a secure hospital, it shortened the time to unsupervised leave by 137 days, a result favouring treatment, rated very low certainty. Dialectical behaviour therapy’s result favouring treatment was a suggestion of fewer self-harm days, from a comparison involving 14 people, also rated very low certainty. No trial of therapeutic community treatment met the review’s criteria.

The National Institute for Health and Care Excellence (NICE), which sets clinical guidance in England, says no drug has established efficacy in treating antisocial personality disorder (ASPD). Its quality standard limits antipsychotic or sedative medication for people with the disorder to short crises, of no more than a week, and to treating other conditions they have.

The Cochrane drug review rated the certainty of evidence as very low for every outcome it examined. The one positive result on the review’s main outcomes is phenytoin, an antiepileptic. It comes from one six-week trial of 60 aggressive male prisoners. Phenytoin was associated with a mean of 0.33 aggressive acts per week, against 0.51 on placebo.

Desipramine showed no difference in social functioning in 29 participants. Nortriptyline showed no difference in global functioning in 20. Bromocriptine showed no difference in global functioning in 18. The bromocriptine trial gave no side-effect figures for its participants with ASPD. But it reported that 12 people had severe side effects, five of whom, all taking bromocriptine, dropped out in the first two days with nausea and flu-like symptoms.

The evidence is strongest in children, and it is not about adults. Parent training programmes reduce conduct problems in children with small to large effects depending on the programme and the measure. A meta-analysis of one widely used programme, Incredible Years, combined the results of 50 studies. It found an effect of 0.27 on disruptive behaviour, averaged over the different people reporting it, parents among them. A meta-analysis of parent-child interaction therapy covered 23 studies and 1,144 participants. It found a standardised mean difference of −0.87 on externalising behaviour. Both figures are standardised, with no units, so that studies measuring behaviour in different ways can be combined.

Callous-unemotional traits, an uncaring and remorseless emotional style, mark a harder subgroup. Since 2013 they have marked a subtype of conduct disorder in the Diagnostic and Statistical Manual of Mental Disorders (DSM-5). A meta-analysis covering 60 studies and 9,405 children found that treatment reduced disruptive behaviour about as much in children with these traits as in those without. But it also found that the children with the traits started with more severe symptoms and finished with more severe symptoms. Treatment had no overall direct effect on the traits themselves; a small effect appeared in some groups of studies, among them those testing parenting-focused components. The authors add that little of the evidence came from randomised trials.

Whether early intervention prevents adult antisocial personality disorder (ASPD) is a separate question, with far less evidence. None of these meta-analyses counted adult ASPD diagnoses; they measured conduct problems and callous-unemotional traits in children. The Fast Track trial did follow children into adulthood. It randomly assigned 891 children with early conduct problems in four US communities, by matched sets of schools, to its programme or to a control group. The programme lasted ten years and combined parent training with social skills training, tutoring and classroom lessons. At 25, 59% of those assigned to the programme had at least one externalising, internalising or substance use problem, against 69% of the rest. Those assigned to the programme were also significantly less likely to meet the criteria for antisocial personality disorder. Fast Track is one trial, of a programme far broader than parent training.

Dropout and age both make improvement in adults hard to credit to treatment.

People with personality disorders often leave treatment early. Across 25 studies of psychosocial treatments, identified by a 2010 review, the median share who did not finish was 37%. One study covered 236 inner-city men in residential drug treatment. Those with antisocial personality disorder (ASPD) who had come voluntarily were more likely to drop out than any other group, including those with ASPD sent by a court. Analyses of only those who finished treatment therefore describe an unrepresentative slice, and court-ordered retention is not the same as engagement.

Donald Black’s 2015 review, citing a US survey published in 1984, says antisocial personality disorder is most common between the ages of 24 and 44. The disorder is less common between 45 and 64, and long follow-ups find many improving decades later. Lee Robins followed up former child guidance patients. Among those who had the disorder as adults, 82 were interviewed about 30 years on. Of them, 12% had fully remitted and 27% had improved, as Black’s review reports them. Black’s own study covered 71 men admitted to the University of Iowa’s psychiatric hospital. It rated 45 of them an average of 29 years after they left hospital. It found 27% remitted and 31% improved. Any study without a comparison group that reports improvement over time is partly measuring age.

A treated group’s drop in aggression can therefore reflect who stayed and the passage of time, before any therapeutic effect is counted.

Nobody knows yet. The two Cochrane reviews of 2020 found too little good evidence to recommend or reject any psychological treatment for antisocial personality disorder (ASPD). About drugs, they found too little to draw any conclusion.

The largest trial came later and randomised 313 men on probation in England and Wales. Adding a year of mentalisation-based treatment cut mean aggression scores at 12 months to 90, against 186 on probation alone. On a diagnostic interview, both groups scored lower for ASPD, the treated men more so. By 24 months the gap in aggression was no longer statistically significant, and over three years the two groups’ offending was about the same. These results come from one trial.

The firmest evidence is in children. Parent training reduces conduct problems, though none of the parent-training meta-analyses counted adult ASPD. One long trial of a broader programme, Fast Track, found fewer psychiatric problems at 25 among the children assigned to it, antisocial personality disorder among them. In adults, two things make any improvement hard to credit to treatment. People with personality disorders often leave treatment early. The disorder is also less common at 45 to 64 than at 24 to 44, so a study without a comparison group partly measures age.

Download the conclusion

The verdict above as one image, with the two numbers that frame it. The numbers show how much lower the treated men’s aggression scores were than those of men on probation alone in the largest trial. At 12 months the gap was 73.5 points. At 24 months it was 25.4, and no longer statistically significant. The image is 1,080 by 1,350 pixels, and anyone may download it and post it.

Download the card (PNG, 1,080 × 1,350)