Monday 22 October 2012

Antipsychotics, autism and core symptoms

NICEly does it @NICE
Analysis of the quite recent publication of the NICE guidance on autism in adulthood in the UK (see here for the full-text) has filled quite a lot of my spare time over these past weeks. That and reading the meta-analysis on medications used in cases of autism in young adults* by Dove and colleagues which concluded that there was "a marked lack of data on use of medication treatments for adolescents and young adults with ASD".

There is quite a bit of material to get through on this latest strand produced by NICE but some interesting information does seem to be included, particularly their advice on the use, or rather non-use, of pharmacotherapy for managing the 'core symptoms' of autism.

Turning for example to page 28 of the document, quite a lot of very direct guidance is offered. So statements like 'Do not use anticonvulsants for the management of core symptoms of autism in adults', don't use special diets for 'the management of core symptoms of autism in adults', don't use oxytocin for 'the management of core symptoms in adults'; and the same goes for testosterone regulation, secretin, HBOT, vitamins and minerals, etc, etc.

The fact that NICE ruled out such interventions for 'the management of core symptoms' might be only five little words but does not necessarily mean that such interventions have been 'ruled out' altogether when looking at other potential features; some of which may have the capacity to impact on core symptom presentation.

Epilepsy is the obvious example - where anticonvulsants have been shown to not only be a life-saver but in a roundabout way also potentially impacting on certain core symptoms in some cases (see Hollander et al**). Indeed bearing in mind the current focus on physical health and autism, one has to perhaps be careful not to 'rule out' say iron supplementation when anaemia may be present, just in the same way that a gluten-free diet would not be ruled out when coeliac (celiac) disease is comorbid. Assuming that is, that such conditions are screened for in the first place. If there is one thing repeated on this blog, it is that a diagnosis of autism is seemingly protective of nothing when it comes to comorbidity.

This got me thinking about the evidence for such medication use in cases of autism and how results were spread between core behaviours and those more peripheral presentations. At the same time I'm mindful of the potential health implications 'around' the use of certain medications such as the antipsychotics (see here for the tragic story of Harry Horne-Roberts) despite some gaps in the knowledge about direct effects.

I've zoomed in on antipsychotics in this post simply because antidepressants, certain classes of antidepressants, have sort of already been covered by past writings (see here). I should at this point reiterate that no medical advice is given or intended from anything in this post. If you need advice, speak to a medical doctor or similarly qualified healthcare professional (preferably not from the Internet 'fountain of information').

Perhaps also at the stage I should re-emphasize the important point made by the NICE guidance on 'managing core symptoms'; that is the triad (dyad!) of behaviours (social, communication, repetitive/restricted repertoire) that make up the diagnostic criteria. This is an important distinction from secondary presentation such as what might be described as 'maladaptive' or 'challenging' behaviours for example tied into the persistence of such behaviours (whether acute or chronic). One has to realise however that secondary behavioural presentations might be able to influence the presentation of core symptoms just as core symptoms may also affect secondary symptoms.

OK after all that, lets see what there is:

  • Well, type in 'autism antipsychotics' and autism into PubMed and you get several hundred results returned. I can't possibly cover every paper in this post as some kind of unofficial meta-analysis, so am instead going to be a bit selective. I am going to try and be objective but you are advised to do your own searches and make your own judgements - don't take my word for it.
  • Discounting a lot of early studies looking at the use of earlier antipsychotics and autism, alongside those pediatric studies, there is quite a body of work suggesting some effect of various antipsychotics on 'peripheral' signs and symptoms linked to cases of autism. So this review by Elbe and Lalani*** (full-text) focused on the use of drugs like risperidone and aripiprazole for managing irritability in cases of autism. They also noted the approval (FDA approval) of risperidone for the treatment of irritability in cases of autism in children and young adults
  • Clozapine, another second-generation antipsychotic (neuroleptic) has also come under scrutiny with regards to peripheral behaviours present in autism. Beherec and colleagues**** discussed the use of clozapine on disruptive / aggressive behaviours; concluding that aside from important side-effects such as weight gain (remembering Harry Horne-Roberts), the drug might offer some benefit to some people. 
  • Stachnik & Nunn-Thompson***** conducted quite a good review of the various data on the use of antipsychotics - second generation antipsychotics - and autism. They concluded that there may be some merit in looking at some of these medications for peripheral symptoms associated with autism but "these drugs do not affect the core symptoms of autistic disorder and are associated with potentially significant adverse effects".
  • Potenza and colleagues****** might to some degree disagree with that last statement about core symptoms with the result of the their study looking at the use of olanzapine and their description of clinical responders showing "improvements in overall symptoms of autism, motor restlessness or hyperactivity, social relatedness, affectual reactions, sensory responses, language usage, self-injurious behavior, aggression, irritability or anger, anxiety, and depression" balanced again with the weight gain issue.

What conclusions could we take from this collected work?

Well, possibly a few. So for example, that the lines between the effects of medication on core and peripheral symptoms are perhaps less clear-cut than one might first imagine. We've kinda seen it before with other classes of pharmacotherapy such as melatonin and its effects on sleep in cases of autism; another peripheral issue not generally noted under 'core' behaviours but potentially impacting quite significantly on the presentation of some core behaviours. 

There is also the issue of cost vs. benefit when it comes to the use of antipsychotic medication in cases of autism. Similar to just about any intervention for autism, benefits must always outweigh costs and costs shouldn't generally be significant in the longer term such as making a person more prone to developing other health issues. Reading between the lines, I am perhaps beginning to realise why NICE made the recommendations they did; based not purely on the efficacy data (i.e. do antipsychotics work in cases of autism?) but rather cumulatively on the efficacy and safety data and the important impact that side effects such as weight gain can have on a person and their long-term health. Indeed The Mental Elf carries quite a good overview of of the research around side-effects from neuroleptics (in children and adolescents).

Finally, and without giving any advice on the matter, more attention should perhaps be given to the use of antipsychotics in cases of autism as being a last resort option. I think back to the area of dementia and that study on the use of painkillers vs. antipsychotics (see here) to decrease levels of 'agitation' and can't help but make a comparison. I'm not insinutaing that all challenging behaviours observed in autism are due to pain nor that there is no place for antipsychotics with appropriate monitoring, but rather that there may be other antecedents to the presentation of challenging behaviours. The trick is to find out what they might be, both at an individual and group level and if needs be, offering good medicines management.

I realise that I might have been a little pedantic in this post; focusing more on wording than the 'spirit' of the text. But given that this and the other strands of NICE guidance are very likely to have a strong influence on the way that autism is viewed and managed by healthcare professionals in the UK and perhaps beyond, as per lots of other areas, a focus on the text is ultimately going to govern what clinical decision are made, or not, about the lives of people with autism.


* Dove D. et al. Medications for adolescents and young adults with autism spectrum disorders: a systematic review. Pediatrics. September 2012.

** Hollander E. et al. Divalproex sodium vs. placebo in the treatment of repetitive behaviours in autism spectrum disorder. International Journal of Neurospychopharmacology. 2006; 9: 209-213.

*** Elbe D. & Lalani Z. Review of the pharmacotherapy of irritability of autism. Journal of the Canadian Academy of Child & Adolescent Psychiatry. 2012; 21: 130-146.

**** Beherec L. et al. Retrospective review of clozapine in the treatment of patients with autism spectrum disorder and severe disruptive behaviors. Journal of Clinical Psychopharmacology. 2011; 31: 341-344.

***** Stachnik JM. & Nunn-Thompson C. Use of atypical antipsychotics in the treatment of autistic disorder. The Annals of Pharmacotherapy. 2007; 41: 626-634.

****** Potenza MN. et al. Olanzapine treatment of children, adolescents, and adults with pervasive developmental disorders: an open-label pilot study. Journal of Clinical Psychopharmacology. 1999; 19: 37-44.


  1. Anticonvulsants -
    I'm told over 80% of ASDs have some sort of seizure activity going on - often sub-clinical - and this can cause hallucinations, which in turn can lead to use of anti-psychotics by professionals who are unaware of problems faced by people with ASDs.

  2. Thanks Janette.

    Certainly hallucinations have been observed in certain types of epilepsy as per the review from Elliot et al:

    and where such behaviours/characteristics are seen, one might need to make some quite detailed evaluation in order to separate them out from other forms of psychosis as being present.

  3. Thank you Paul - this is actually extremely helpful in our personal situation and set me off on a paper trail of equally informative articles!


Note: only a member of this blog may post a comment.