Autism Plus Psychosis Equals Mass Murder?

By Neuroskeptic | June 27, 2013 4:37 pm

According to a new paper, mass shootings such as Sandy Hook and Aurora may be the result of Autism plus psychosis: A ‘one-two punch’ risk for tragic violence?

The first thing to note about this paper is that it’s in Medical Hypotheses.

I don’t normally take seriously anything that appears in this rather unique journal. This paper is, however, co-authored by Edward Shorter, an eminent historian of psychiatry; his book about it was a big influence on me. So it deserves a fair hearing.

The authors set the scene:

In the recent series of mass murders in Connecticut, Colorado, Norway and elsewhere, a pattern seems to be emerging: young men whose social isolation is so extreme as to verge on autism apparently become prey to psychotic ideation. And under the influence of this ideation they wreak terrible violence.

…What is actually the matter with these young men and how should we as a society conceive their pathology?

The answer, we’re told, is a combination of autism, and psychosis. Autism is not associated with violence per se, but psychosis is – and rates of psychosis are higher in autistics. What’s worse, in such cases, psychotic symptoms may go undiagnosed and untreated because they’re written off as just part of the autism.

The authors offer a remarkable analogy to illustrate their point about how an illness on top of an existing disability can lead to tragedy:

A middle-aged woman with severe mental retardation was brought to the ENT service for debulking of a metastatic melanoma of the nose approximately the size of a peach. The tumor had been diagnosed only a few days earlier; yet it had already invaded the brain, leading to a terminal condition.

The surgical resident commented on how unfortunate it was that the melanoma wasn’t detected earlier when it could have been easily ablated… Another team member speculated this may often be the fate of individuals with disability; few people care, and no one notices until it is too late

The trouble is, this is, as the authors admit, all speculation. The authors write on the very first page:

It is unknown whether [these killers] had an autism spectrum disorder (ASD) or other neurodevelopmental condition; similar conjectures regarding psychotic illness and personality disorder remain unanswered.

This strikes me as a fairly big limitation. But even if we grant that they did all have autism plus psychosis, what would that mean? I don’t think it would explain much. We all know there was ‘something wrong with’ these individuals, they fit a very definite ‘type’, but I’m not convinced that putting a label on it (or two) helps us to understand it.

ResearchBlogging.orgWachtel LE, & Shorter E (2013). Autism plus psychosis: A ‘one-two punch’ risk for tragic violence? Medical hypotheses PMID: 23786904

  • tanya

    If you know anything about psychosis very few people with psychosis get violent. ppl with autism do.

    • Nitric-X

      Bold statement, tanya. But is it backed up data? Actually, no. Indeed, patients with schizophrenia (I’d rather use this term, still bad, but better than psychosis which has a myriad definition) contribute little to violent acts on the overall population level. However, the relative risk to behave violently is increased as compared to healthy controls (see e.g. this open access paper: – the OR for homicide in scz as comp to hc is 19!).
      And as you throw in personal experience – “if you know anything about psychosis” – well I do, having worked on a closed ward for 5 years, and yes, psychotic people can be highly violent. i got bitten twice – which is little damage done considering the number of patients during this time.

  • psycritic

    I wonder how the authors would explain the knife attacks in Chinese schools, which are mostly done by middle-aged men instead of young men having their initial psychotic breaks.

    Looking at societal violence without examining the society in question seems to be a very limited perspective.

  • Grizzled_Stranger

    On the record; Nancy Lanza searched desperately for adequate medical care for her son. James Holmes’ psychiatrist reported Holmes to the police as a danger to himself and others. Jared Loughner repeatedly came into contact with Pima County, Arizona deputies, who described his highly erratic actions as “crazy.” Less notoriously, Sikh Temple Killer Wade Page had a record of psychological disturbance. And of course the list of “active killers” who should have gotten mental health care up to and including commitment goes on and on and on.

    On the record, more than eight out of ten “active killers” should have been committed long before they became mass murderers. Yet nothing more than “words and bandaids” was done for any of them.

    Since we only have a few of the mass murderers to investigate; we really cannot prove autism plus psychosis is the problem. But at least that hypothesis is not as absurd as the much touted “trigger pulls the finger” theory.

    • Jon Nixon

      Good idea, just commit the crazy folks….. the problem is, that until they break the law, or it can be proven beyond a reasonable doubt that they are an eminent threat to themselves or others, they remain free persons in a free society…and free to exercise their second amendment rights as law-abiding citizens. Give matches to a five year old , don’t be surprised when the house burns down…….,and only a fool would blame the matches.

    • Neuroskeptic

      Crazy and erratic and disturbed – absolutely. In most cases.

      But the question (here on this post) is, were they psychotic and autistic? That’s less clear.

      • de Broglie

        Why do African-Americans cause so many murders in the United States? Do they have higher levels of autism. I am not from the US, so I do not know.

      • Neuroskeptic

        I remind commenters to please stay both on topic and on their rockers.

    • wylekat

      It’s just an easy out. I think it’s much more than a little psychosis- and could be dealt with using real world solutions, and not meds, in most cases.

    • andrew oh-willeke

      Imagine a simple, statistically validated risk scoring instrument with twenty questions or so that would meaningfully separate people in contact with the mental health system who posed a high risk requiring immediate triage prioritization from those who might be clearly in need of mental health care but were not at high risk of being mass killers. If it excluded 80% of the people in the risk pile, even if it was massively overinclusive, that would be a huge deal. A call to police about someone who is a danger to himself and others v. a call about someone who scored in the likely mass killer range on a standardized profile would have a different effect.

  • Zachary Stansfield

    So historians are trying to map the fuzzy categories of psychiatric illness onto the even fuzzier categories of criminal tendency…

    This seems like a work of futility.

    • Wouter

      I don’t think this is a really fair remark. Although one might wonder how succesful and correct this theory is, they’re trying to postulate signs that will help recognize potentially violent individuals, before they actually turn to violence. Doesn’t sound that futile to me.

      • Zachary Stansfield

        “Doesn’t sound that futile to me.”

        It sounds as though you believe that the overall goal is commendable, and itself not necessarily futile. I won’t disagree with that idea, in principle.

        However, I am trying to point out that the method they are using (e.g. relying upon broad, heterogeneous categories with only moderate levels reliability and validity) will never achieve this aim (e.g. predicting an extremely rare outcome with high specificity). This argument can be made relatively simply: precise, long-term prediction requires a much higher standard than simple categorization.

        I don’t have the numbers in front of me, but there is undoubtedly a technical argument to be made here: in order to have sufficient predictive validity to identify rare outcomes with extreme precision (high specificity and sensitivity), then our models (e.g. mental illness categories) must be extraordinarily precise as well. They aren’t.

        • andrew oh-willeke

          We don’t need extreme precision. A model that that has a 10,000-1 false positive rate, and a 25% false negative rate would be a profound improvement on the status quo that would probably cut the number of mass killings in the developed world by 50%.
          Fundamentally, the key issue is how distinct people who are at a high risk of committing a mass killing are from people who have a negligible risk of doing so. For example, if just 5% of mass killers do not suffer from a psychosis, and just 1% of the population suffers from psychosis and we can diagnosis psychosis with a 100% false positive rate and a 25% false negative rate, then psychosis is a pretty good first order screen for mass killers (but also insufficient). If 95% of mass killers are men and misdiagnosis of male gender for potential mass killers is a one in 5,000 event (have there ever been any gender ambiguous mass killers in all of recorded history?), then this is a pretty good screen. If broad spectrum autism traits are found in 90% of mass killers and in 5% of the general male population and this is not extremely comorbid with psychosis, then this is a pretty good screen. Throw in some characteristic environmental stress factors, some perpetrator age related factors, some factors to exclude people whose physical health or intellectual capacity isn’t great enough to pull something like a mass killing off, and an absence of a sustained period of institutionalization (that seems to be a common thread) and you’re getting somewhere.

        • Bad Pit

          Violent behavior can stem from truama as a child. Abusive parents, for example.

  • JT

    Better to stick with the data, I think. More convincing too (h/t Michelle Dawson on Twitter):

    • Grizzled_Stranger

      I hate to say this but I wasted forty bucks on that study.

  • wylekat

    What utter psychobabble BS. The REAL problem is the lack of help when someone asks or even pleads for it in desperation. No legal recourse. Being marginalized for any number of petty reasons. Being ignored when one easily and coherently spells out the problem. Problems like depression having the GD *symptoms* dealt with, when the problem should be tackled, COULD be tackled, and is left to just keep generating symptoms (and therefore money for pharma).

    What happens after an attack is in my opinion, is even uglier. The media spin put on it to make sure everyone is assured said person was just ‘unhinged’ to begin with. The unbelievable focus on the victims (to a point of insanity), some ‘feel good’ legislation- and another lost soul takes a trip into the pit. Or, just killed and swept under the rug.

    I am personally seeing such stuff in action.

    • Grizzled_Stranger

      I drink coffee with a psychologist and hear all too much about “no place to get long term help for my patients.” I talk to a couple of psychiatrists who left California because the legislature can fund all sorts of things, but the mental health people must replace a burned out light bulb out of their own pocket.

      In the grand scheme of things the fifty or so victims of “active killers” are small potatoes compared to the 9,507 criminal on criminal murders. But the ConC murders have declined by more than half in 18 years as potential victims have armed themselves.

      Literally nothing is done for the mental cases like Adam Lanza, and the rest. And that is a shame and a disgrace

    • Search On

      Thank you! Without minimizing the tragic loss of life, what good is some generalized hypothetical check list to assess the potential for such horrible acts if there is no effective support, treatment, or intervention available before it is too late? As was stated earlier, the signs of trouble were there in most cases and the people/agencies/authorities that could have intervened apparently did not. The reactive sensationalism in the media and armchair diagnoses that it perpetuates does little except imply that “they” (pick your diagnosis) are to blame.

  • Buddy199

    It is unknown whether [these killers] had an autism spectrum disorder (ASD) or other neurodevelopmental condition; similar conjectures regarding psychotic illness and personality disorder remain unanswered.

    Then why publish a paper? They could just as well have been under the influence of the full moon or evil spirits.

    • Neuroskeptic

      A paper about the full moon affecting mass murders would probably get into Medical Hypotheses. Evil spirits… maybe too much even for them.

    • andrew oh-willeke

      A journal like Medical Hypotheses provides a forum for credible, smart professionals who aren’t in a position to carry out research themselves (e.g. due to lack of access to people in correctional custody and law enforcement case files) to make constructive suggestions to the entire class of people (some unknown to them) who may be in a position to facilitate these kinds of studies (e.g. an Interpol or UN agency research division that has understanding’s with national health care databases in member countries that allow that big data to be used for research purposes in ways it couldn’t be in the U.S.).

  • jones

    and what about the 50 thousand other murders that happen every year in the usa ?

    • Grizzled_Stranger

      Well, Jones, someone has played a joke on you. There are countries with 50,000 murders a year; China, Russia, Venezuela and Brazil, for four. But the United States had only 14,612 HOMICIDES; of which 12,664 were murders and only 8,583 were gun related. That is virtually a 50 percent decline in the homicide rate since the peak of gun control fever in 1993.

      Look up “2011 Crime Victimization Survey” for the official data, not that off the top of someone’s head.

  • Pingback: Autism Plus Psychosis Equals Mass Murder? | Follow Me Here...()

  • Lewis Ambrose

    Have to be very skeptical of this kind of open speculation. 1) because as you noted, it is all speculation, and 2) there is a strong potential for this kind of speculation to be misunderstood and misused.

    Some problems with this hypothesis– a) diagnosing autism in an adult with schizophrenia is far from straight-forward, and more importantly, b) even if they have identified a genuine risk factor, we have to consider the rate of autism and the rate of schizophrenia in the general pop., the overlap being a sizable population in its own right, of which what proportion become mass slayers of children? We need to that number to think about this rationally.

  • petrossa

    It’s just as idiotic as diagnosing historic people with whatever mental condition is in fashion, as idiotic as retroactively predicting the weather wrong, as idiotic as thinking that if you slap a label on something it suddenly becomes something else then it was before.

    Psychiatry/psychology needs to be degraded to ‘beliefsystem science’ on the level of astrology, homeopathy etc.

    Well at least they take themselves immensely serious, otherwise not many would.

    • Nitric-X

      “Psychiatry/psychology needs to be degraded to ‘beliefsystem science’ on the level of astrology, homeopathy etc.”

      Petrossa, most of your points are usually polemic but convey some points of truth which are not well liked by most neuroscientists. However here you’re way too far out. Psychiatry/psychology are real sciences, they put up hypotheses which can be tested, and this is exactly what happens. Whereas astrology/homeopathy usually don’t – if it’s done, hypotheses are mainly refuted. Plenty of blogs on pseudosciences are out there, many of them are excellent, and none of them psychiatry/psychology in. You are not doing research a favour with such blatant propaganda – if all neurosciences are would be bogus, as you state, then you have to come up with an alternative explanation that is backed up by evidence and can withstand Occam’s razor. There you go – challenge taken?

      • petrossa

        Psychiatry is not a neuroscience. In fact i am very much in favor of neuroscience setting the norm scientifically what is a disorder and what not based on mal/different functioning neural networks.
        Psychiatry/psychology are a belief system, a circular logic based labeling mechanism with no known real cures to its name as neuroscience does have. Brainsurgery does have clear curable effects, medication by psychiatrists have only symptom suppressing effects. Psychotherapy is the ultimate placebo, it works when it works else it doesn’t.

        • Nitric-X

          Petrossa, obviously there are some misconceptions here, and as they are kind of on topic I’ll elaborate.

          First, you keep on saying the same over and over again without either backing it up or offering alternative explanations. This merely as a side comment.

          The most important thing is that obviously you don’t get the very idea of medicine and medical research. What can be considered an illness is always a man-made decision, or rather societal agreement, and thus “top bottom” – it can never be “bottom up”, as nature clearly does not “think” in concepts like disease or health; if at all, nature cares about reproductive success which is not at all the same. What do I mean by this? Well, most people would agree that mycordial infarction is a disease. And most would agree that coronary heart disease is just the same. Now we all know that there are genetic risk factors to it, as well as e.g. hypertension or increased blood lipids. Usually we thus consider them diseases too, which also makes sense, b/o treating them prevents MI. However where we draw the the line is entirely based a) on epidemiological studies and b) consensus, it does not derive from any molecular etc. study. Similar arguments can be made for prostate cancer etc. etc.: where we draw the line is often defined rather than evident as such. In some cases it is fairly easy to draw the lines, in other cases it is not. Either way the clue is that “disease” is a man-made concept and not a natural law. (you might argue that e.g. for monogenetic disorders the distinction is very clear. Yet then again, to consider the phenotype a disease is man-made, nature only cares about reproductive success. A nice example would be sickle cell anemia, considered to be a disease but yet quite advantageous in malaria regions.). Nature is not binary, it is about variation, and from when on we call it “disease” is a human decision and according consensus within societies. Western societies these days base their definition mainly on “individual suffering” (rightly so, in my opinion), whereas not so long ago “norm deviation” was more the route which was followed. And other concepts are possible – e.g. whether or not one can maintain his/her role as a worker or whatever. Just to make it clear once again, “disease” is an entirely human concept (e.g. when we talk about a bacterial infection our whole concept of disease is quite anthropocentric, no?).

          Having said this, it becomes pretty evident that disease will never be defined bottom-up by any scientific means. For instance, assuming you study gene ZEBRA (entirely fictional). You find out it is expressed in melanocytes and regulates cell division and orientation. A mutation in ZEBRA causes abnormal skin pigmentation patterns, so that the individual displays by funny black stripes. Does this mean it is a disease? Molecular biology is agnostic about this. If the mutation also goes along with decreased life expectancy due to cancer, we’d say yes. If it goes along with resistance to melanoma, longevity, and the ability to have unlimited New York Cheesecake without gaining weight we’d probably say no (and all want to have it). Again, whether or not it’s considered an illness is man-made.

          So the medical-scientifical process is recursive and adaptive. Man defines whatever illness derived from symptoms / syndroms, has an idea about pathophysiology, designs respective experiments, which then again can inform the idea about the illness per se – it might be confirmed, changed, whatever, in order to refine diagnosis and hopefully treatment. And then back and forth again.

          Psychiatry by no means is different from other medical sciences in this respect, probably only a bit slower (IMHO due to the fact the brain is a) quite complex and b) hidden in the skull). Every medical speciality started to define disease complexes which where then studied scientifically, and we saw (and still see) many surprises on the way. Plainly it makes me sick as hell that psychiatry is always considered to be different. IT IS NOT. It may not yet be where oncology is, but then, look at oncology a mere 30 years ago. Psychiatry is clinical neuroscience, with testable hypotheses and disease concepts, and if you call it a belief system a third time without backing it up it doesn’t help a lot: psychiatric disorders are disorders by human definition just as much myocardial infarction is. Only that we have much better biomarkers for the latter.

          So I hope that I made clear why neuroscience can never set any norm. It can provide pathophysiology and etiology, not more, not less, but a normative act can only ever be made by a human being (in other words – assume a Gaussian distrubution, where would be the norm?)
          Next… you always lump psychiatry and psychology… why, actually? Those are very, very different things, and psychology for most of the time isn’t about pathology at all. On the other hand – neuroscience and cure? Really? Which cures do you specifically refer to?
          Further… “Brain surgery does have clear CURABLE effects, medications […] only symptom supressing effects”. While the latter statement is partly true, I question its relevance (so if it’s supressing symptoms, why bother?) and furthermore, what does it actually mean? If a depression remits 2 months earlier, does this mean symptom suppression? Cure? Again, is this a natural phenomenon? Or rather definition? And who the f*** cares, if this treatment results in effective reduction of suicide rates = deaths = decrease in mortality? (which it does, by the way). We only have these kinds of weird discussions in psychiatry, in my time working on a internal ICU I have never heard anybody complain that post-MI treatment would be “only symptom suppressing”. Oh yes, and the statement on brain surgery in this context is so totally wrong that I doubt you know what you are speaking of. Just shut off a DBS device against Parkinson’s and you’ll see that this indeed is “only” symptomatic. (if you’re talking about brain surgery against tumors we talk about totally different things here. Probably you could cure addiction pretty completely by cutting out the N. Accumbens, but I doubt somebody would volunteer…).
          Finally … “psychotherapy is the ultimate placebo” (etc.): either you are ignorant or illiterate about the literature here. There are plenty studies out there (randomized, controlled, etc.) that clearly show e.g. the effect of CBT in OCD or panic disorder. Especially the latter statement really made me think whether you have at all any knowledge on the topics you comment upon?

          • petrossa

            Thank you for your extensive and well meant explanation. However the misconception is on your side. Psychiatry is not a science. Since someone went into that in detail here i won’t repeat it all but just the catchphrase: , “whereas psychiatric classification is based on the subjective perceptions and votes of psychiatrists.”

          • Nitric-X

            Petrossa, neither this is an answer to my post, nor did you get my main point. I guess at this stage it makes no sense discussing further.

          • petrossa

            Evidently this goes to far over your head, although i though i’d explained myself clearly. Indeed there is no point in discussing this any further.

  • Grumpy Marmoset

    Wachtel (Kennedy Krieger Institute) has the most unusual, non-mainstream beliefs about autism in the first place, such as clinging on to the diagnosis of “catatonia”, as well as treating autistic children and adults with a mixture of antipsychotics and electroconvulsive therapy – ECT is not recommended anywhere as a “treatment” for autism, except by Wachtel and a few close colleagues.

    Wachtel, L. E., & Dhossche, D. M. (2010). Self-injury in autism as an alternate sign of catatonia: Implications for electroconvulsive therapy. Medical Hypotheses, 75(1), 111–114. doi:10.1016/j.mehy.2010.02.001
    Wachtel, L. E., Dhossche, D. M., & Kellner, C. H. (2011). When is electroconvulsive therapy appropriate for children and adolescents? Medical Hypotheses, 76(3), 395–399. doi:10.1016/j.mehy.2010.11.001
    Wachtel, L. E., Dhossche, D. M., Reti, I. M., & Hughes-Wheatland, R. (2012). Stability of Intellectual Functioning During Maintenance Electroconvulsive Therapy. Pediatric Neurology, 47(3), 219–221. doi:10.1016/j.pediatrneurol.2012.05.012
    Wachtel, L. E., Hermida, A., & Dhossche, D. M. (2010). Maintenance electroconvulsive therapy in autistic catatonia: A case series review. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 34(4), 581–587. doi:10.1016/j.pnpbp.2010.03.012
    Wachtel, L. E., Reti, I. M., Dhossche, D. M., Slomine, B. S., & Sanz, J. (2011). Stability of neuropsychological testing during two years of maintenance electroconvulsive therapy in an autistic man. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 35(1), 301–302. doi:10.1016/j.pnpbp.2010.11.013

  • Eric Cartman

    Certainly there are killers with ASD, like Lanza and Kaczynski. But claiming someone on the schizo- spectrum is autistic because of “social isolation” is incorrect, because social isolation is a negative symptom of schizophrenia and schizotypal PD.

    • Pauly Colland

      I think whether someone is sociopathic is the most important thing regarding the potential for mass murder. Whether they be schizophrenic or autistic. Once psychosis emerges in this spectrum , sociopathic tendencies could push them over the edge.

  • Pingback: Autism Plus Psychosis Equals Mass Murder? : Neuroskeptic | Apparel straight()

  • Pingback: Linn Plus Berry: A Shareholder Friendly Deal – Seeking Alpha | Clothes fidelity()

  • andrew oh-willeke

    “We all know there was ‘something wrong with’ these individuals, they fit a very definite ‘type’, but I’m not convinced that putting a label on it (or two) helps us to understand it.”

    One of the characteristic features of mass killings is that while mental health issues are clearly a key factor that links perpetrators that no one has come up with a very specific, operational definition of people who are “at risk.” Existing law uses the threshold of involuntary psychiatric commitment (typically ordered for 72 hours by a mental health profession based upon perceived threat to oneself or others, sometimes extended upon further evaluation) to bar people from purchasing firearms, a threshold that includes large numbers of people at low risk for mass shootings (e.g. dangerously extreme anorexics), but fails to include many people who pose a serious risk. And, any measure based on prior criminal conduct is both wildly overinclusive (since many criminals are sane and not violent), and wildly underinclusive (since many people who commit mass killings have little or no criminal record prior to their final incident).

    From a public safety and law enforcement perspective, a risk profile that captures the 1% of the population that accounts for most of the risk (which psychosis would) includes a difficult to afford heightened scrutiny 3,300,000 million people or so, but with a risk profile that captures, for example, just 5% of the people in that 1% (about 180,000 people) (i.e. an estimated share of autistic people with psychosis), it is much more feasible to put in place an early warning supervision and monitoring that would be effective enough to prevent some (certainly not all) mass killings. I’m not sure that this particularly analysis, while plausible, is actually correct, but the difference between a 99% accurate screen and a 99.9% accurate screen is immense in terms of its real world usefulness.

  • Pauly Colland

    Ok, I hesitate to use words like empathy and morality , because people have the right to behave the way they want as long as they live within the law , but lets say people , who are devoid of knowledge of what it means to be human and live within humanity , then yes , with psychosis , they could potentially be dangerous. Short rule of tumb , Don’t isolate yourself and know what’s going on in the world , read the news etc.

  • Pingback: Autism Plus Psychosis Equals Mass Murder? : Neuroskeptic | National jeans()

  • Pingback: Gay Talese’s Daily Routine, Plus a Money-Saving Tip from the Godfather of Literary Journalism | National jeans()

  • Pingback: Extra Slackery News Tidbits (Plus Shameless Plug): July 17, 2013 | Clothes()

  • Pingback: Autism Plus Psychosis Equals Mass Murder? – Neuroskeptic | Attire stuff()

  • Maria Taheny

    insecure avoidant or disorganized attachment, loss and separation of a parental figure (the father of Adam Lanza divorced his mother) potential change of home situation (his mom was planning a move to North Carolina) all can trigger anxiety in a child with autism. Studies show the anxiety is worse in higher functioning children with autism. Adam was diagnosed with Asperger’s so the possibility that he was going over the edge is evidenced in the behavior at home. Bottom line is parents need greater supports to know how to spot warning signs (psychosis develops more often during puberty in autism cases) and how to know when their child is struggling. Adam’s mom never decorated the house for Christmas, never went into his room to clean and only spoke to him by email, even though he lived in the same house, according to the police report. This speaks to a rigid routine of familiarity in Adam, and this is a big warning sign. The risk of Mom moving the family to NC, combined with the loss of his father and his inability to endure even the smallest changes in the home give more support to a child who was struggling in living. That is where mental health could help.

  • Dennis Embry

    Speculation is not idle when supported by reasonable epidemiological data and case examples. Consider some important epidemiological data. 1) First episode psychosis is absolutely associated with higher rates of homicide across multiple countries Please see the paper in Schizophrenia Bulletin entitled, “Rates of Homicide During the First Episode of Psychosis and After Treatment: A Systematic Review and Meta-analysis.” After treatment, folks with schizophrenia have a lower rate of homicides—but not with first episode (before treatment). Jared Lougher (Tucson) and James Holmes (Aurora) a perfect case examples of first episode psychosis. First episode psychosis is associated with a 15-fold increase in homicide rates; after treatment, which typically occurs after such first-episodes, rates of homicide drop below the base rate of the population. This accounts for the paradox in epidemiological findings.

    The Archives of General Psychiatry (Feb, 2010) published the first randomized control, longitudinal study to avert first episode psychosis, which had dramatic effects on reducing such psychoses. Both myself and a premier scientist with the National Institute of Health testified about this science at Congressional Field Hearings recently in Connecticut. The mechanisms can be well mapped out in brain-scans as well as by epigenetic mechanisms, again most of which is available for folks to read at the National Library of Medicine aka www pubmed .gov (sorry have to type this oddly or the DISCUS will prevent the posting of a hyper-link).

    The symptoms of autism are normally evident in early childhood, even for children with “high-functioning” autism apparently like Adam Lanza. First episode psychosis, however, happens in late adolescence and early adulthood for the most part—as clearly evidenced by Jared Laughner and James Holmes.

    First episode psychosis can be significantly increased by a whole array of causes, which most people in the public don’t know unless they spend time reading through high-quality scientific journals. What is also clear is that the neural pathways associated with autism do overlap with mental illnesses like schizophrenia. We know that some factors can worsen the risk of first episode psychosis such as heavy marijuana use during adolescence (true of Loughner), and we know that sleep deprivation/loss significantly increases risk for first episode psychosis. Vitamin D deficiency (D3) is significantly implicated in schizophrenia in later life, and implicated in autism. The mechanisms of that are quite interesting.

    The autopsy of Adam Lanza could not yield much about bio-markers, as many of those biomarkers degrade rapidly. His body was removed many hours later. Since he was not seen professionally, we don’t have medical records for forensic examination.

    But if you look at the topography of his behaviors as best we know and just his physical visage from photos, his expressions are congruent with individuals who are prodromal (the fancy word for showing symptoms) for first episode psychosis. If that did occur on top of the documented high-functioning autism, his case would have been quite difficult.

    As a good prevention scientist and experienced with psychiatric hospitals, I’d be willing to bet in Vegas that Adam Lanza did have a first episode psychosis on top of his autism. It’s one of the most consilient, concise explanations with good science.

    There is another message that is subtle in this instance. Here was was well-to-do family, albeit like most families with some “issues.” They were not poor. Apparently, neither parent seemed to grasp what to do or who seek for assistance in one of the richest communities in America. Well before this event, I had been involved in planning a major prevention initiative in Connecticut to provide universal supports for parents/families facing mental, emotional, and behavioral challenges. The plan was for three communities, but funding short-fall cut that down to two communities. The area of the shooting was not selected because of not enough funds for the programmatic prevention.

    A lesson to recall is that all families need support for practical strategies to avert serious issues like this. All of this becomes more worrisome when one carefully examines the epidemiological data on mental, emotional, and behavioral disorders among America’s young people: Such disorders are steadily increasing, as documented by the 2009 Institute of Medicine Report on the Prevention of Mental, Emotional, and Behavioral Disorders Among Young People. The tragedy is that the US has amazed the best science in the world to prevent such disorders, but we don’t use the science. Other rich democracies are, but the United States.

  • AutismAcceptanceDay

    have looked at the actual study online, and using “data” from the
    study, some of the main sources of which are the UK tabloid “Daily Mail”
    and a site called “murdermedia,” I have found that there is no
    correlation, either in the study or anywhere else, between autism, brain
    injury, and mass murder. In fact, the study was not able to find any
    correlation between autism and brain injury. In addition, the literature
    review shores up its claims by bringing in unrelated issues, such as
    adoption studies, and the six “definite diagnosis” offenders are not
    definite, as shown by a search I did which includes psychological
    examination of offenders.

    I have
    written about the many errors of both fact and speculation in a blog
    post here, and science writer Emily Willingham has also written about
    the “study.” My post debunking this current, flawed, attempt to link Autistics and murder is:

  • Brad Pitt

    People with mental illness are supportive and nurturing, especially to one another. Psychopaths should definitely not be categorized as mentally ill. Doing so worsens the struggles of the mentally ill when all they want is love and support. I don’t know enough about Autism to comment.

  • JMK

    Sensationalist crap like this is the reason autistic folk are institutionalized, dehumanized and abused on a daily basis.
    Then you people turn around and say it’s us who lack empathy and humanity.

  • Oli D. Nejad

    I have an Autism Spectrum Disorder as well as a past diagnosis of
    Psychosis and (postulated) Schizophrenia. I hate to rain on the fear parade here folks, but at no point throughout the
    course of my internment in the psychiatric setting did I ever display
    any homicidal ideation, or aggressive behaviour with intent to harm.

    All of the other assumptions made by this study aside, psychosis alone is tremendously misunderstood and wrongly cross-connoted with violence. You can have a non-aggressive psychosis, as I did.

    It’s a bit like alcohol, if aggressive in nature, chances are you’ll be an aggressive drunk. The biggest predictor of violence is… violence. Not mental illness or ASDs.

    A psychiatric historian (emphasis on historian) offering blanket
    suppositions after drawing links as tenuous as those that influenced and constituted my delusions, using an infantile psychiatric history of
    barbarism as its basis, doesn’t strike me as a watertight foundation for
    unbiased critical analysis.

  • Just call me Joe

    Fast forward to 2015, and we have another autistic with Asperger’s who went on a murderous rampage at a college in Oregon.

    It is becoming harder and harder to deny any link between autism and murderous psychopaths.



No brain. No gain.

About Neuroskeptic

Neuroskeptic is a British neuroscientist who takes a skeptical look at his own field, and beyond. His blog offers a look at the latest developments in neuroscience, psychiatry and psychology through a critical lens.


See More

@Neuro_Skeptic on Twitter


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Collapse bottom bar