Sunday, May 31, 2009

Does head injury cause schizophrenia?

Dear colleagues,

Welcome to my first blog on neuropsychiatry. For this, I have chosen a topic which I have had to confront recently from a medico-legal viewpoint. There have been a number of competent reviews of this topic, but somehow clarity has not been achieved. while not attempting to review the topic once again, I present some reasons why a causal relationship between schizophrenia and traumatic brain injury is difficult to establish. While this does not pose a major problem for Psychiatry, it makes lawyers trying to argue for their clients quite nervous.

I think this is an extremely difficult question to answer as it is bedevilled by numerous methodological issues. It is important to understand the methodological issues in order to fully appreciate the reasons for the differing opinions between experts:
1. Schizophrenia is a very heterogeneous disorder, and may in fact be a group of disorders. The question then arises: are we looking for an association with a particular subtype of schizophrenia, or schizophrenia in general, or indeed any psychosis that resembles schizophrenia?
2. Head injury is also extremely varied, and may vary from a whiplash injury with no loss of consciousness to extensive brain damage and bleeding. The brain regions affected can vary considerably depending upon whether it is a closed or open injury, caused by deceleration or a missile, etc, and a number of other factors, including associated injuries and the psychological aspects of the injury. Not all head injuries come to medical attention.
3. In the face of this heterogeneity, establishing cause is daunting at the very outset. However, one can try to narrow this task down by clearly defining both schizophrenia and head injury, and this is the first place where studies differ and therefore produce differing results.
4. The process of establishing a cause of a disorder in medicine is complex. The first step is an epidemiological one, in which investigators establish an association (in this case, for argument, that schizophrenia and head injury have more than a chance association). Much of the debate in the literature is focused on this (I will come to this later). Of course, it is implicitly acknowledged that head injury would occur first and schizophrenia later for head injury to be causative. However, the reverse association is also possible – individuals with schizophrenia might be more vulnerable to head injury because of their proneness to accidents or violence or due to postural or gait problems due to the drugs. This is usually not an issue if schizophrenia has been diagnosed prior to the head injury (although the issue of worsening of psychosis can then be raised). It is more problematic if an individual who suffers a head injury is in the prodromal or prepsychotic phase of schizophrenia, which is possible since schizophrenia may develop very slowly over months and sometimes years. There is also a suggestion that individuals prone to schizophrenia may have some abnormalities much before they become unwell, and these could be present from birth. These abnormalities could also make them vulnerable to accidents and head injury.
5. Even if an association is established, it does not mean that a cause had been found. Epidemiological studies have reported an association between head injury and a number of neuropsychiatric disorders, which include dementia, Parkinson’s disease, schizophrenia and bipolar disorder, among others. There are many reasons why spurious associations can be thrown up by epidemiological studies (ascertainment bias, recall bias, etc.). An association must be followed by biological plausibility of the cause, and an understanding of the mechanism by which the putative cause produces the result. This is where the major difficulty with head injury and schizophrenia lies. Since we do not know the cause of schizophrenia, no association can be dismissed outright as being implausible if the brain has been affected in some way, however minor. For this reason, no head injury can be dismissed as being irrelevant if it is thought that brain trauma has occurred.
6. Furthermore, there are different types of causes. Of course, head injury cannot be a necessary cause of schizophrenia. Is it however a sufficient cause, or should it occur on a substrate of vulnerability for schizophrenia to occur? Since head injury is very common, and schizophrenia is an uncommon disorder, most models propose that it is unlikely to be a sufficient cause. It therefore is a contributory cause, if at all. It is proposed as one of multiple possible factors that increase the risk of developing schizophrenia, of which genetic factors are the strongest. This topic has been extensively reviewed in many places, and the Tandon et al (2008) article is a good recent summary. This paper does show that of the many factors, head injury does not feature high on the list.
7. But what about the epidemiological evidence? That experts have reached differing conclusions suggests that the message is not clear-cut. This is partly because medical research has a different objective from a legal discourse. In the case of schizophrenia, the objective of pursuing this association, minor as it might be, is to develop some insights into the mechanisms by which schizophrenia develops; it is not intended to apportion blame. Every study has its limitations. The ideal study would be one in which a very large cohort of individuals who sustain a head injury, and a comparable control group without a head injury, are followed up over many years to determine whether the rates of schizophrenia are different. Even such a massive study cannot be fool-proof as many other factors will need to be considered, and totally unbiased populations do not exist in reality. The question, whether head injury causes schizophrenia can therefore, in my opinion, cannot be answered until we understand the pathological mechanisms of schizophrenia itself. There are many such debates in the medical literature involving complex disorders, the mechanisms of which are not known.
9. What is the consensus? I would agree with the recent review of Dr Edward Kim in the journal Current Opinion in Psychiatry in 2008 (I invited Dr Kim to do this review as editor of the Neuropsychiatry section of the journal). The consensus, in my opinion, is that we do not know for sure whether head injury does increase the risk of schizophrenia. If it does, the relative risk is most likely to be in the order of 1.5 to 2.0. I agree that the recent Scandinavian studies (Nielsen et al, 2002 and Harrison et al, 2006) are the most systematic epidemiological studies, but they are not without their limitations. We should place this relative risk in context. The risk of an individual developing schizophrenia in their lifetime is about 0.7% (figures between 0.5% and 1.0% are cited). Having received a head injury possibly increases this risk to 1.0 – 1.4%, according to these estimates. How this risk estimate translates into the occurrence of schizophrenia in one individual is, however, not known, but the conclusion would be that head injury would be a minor factor, if at all. There is no consensus on the type of injury and this risk, although some evidence suggests that more severe injury may be worse. It is also not established whether injury to any particular brain region is worse for this risk, although temporal lobe injury has been suggested by some studies.

How can one reach a decision about causal association in an individual case?

From a medical perspective, this is impossible. It is therefore more of a legal issue, and stems from the fact that the law is unable to deal with probabilities and possibilities in the same manner as medical science can. The difference between the two perspectives could be narrowed if legal argument was able to apportion part-causation (or a contribution from a risk) and judge responsibility or blame accordingly.

Having said this, clinicians do sometimes encounter individuals who never recover following brain trauma and go on to develop what appears to be schizophrenia. It is quite reasonable for the clinician to then ask the question whether the trauma did in fact cause the disorder. Family members are also impressed by the fact that the individual was well prior to the head injury, and they therefore blame the head injury for the problem. In fact, their retrospective recall of the individual’s functional status is often biased by this knowledge, and they usually paint a positive picture of the pre-injury individual. We know from many other examples in medicine that these inferences can be erroneous, and that is why systematic evidence is collected before an association can become an established fact. In the case of head injury and schizophrenia, the association cannot be said to be established fact, although it cannot be dismissed outright either.

What is then a reasonable way forward? For medical science, it is more research on many fronts, including better epidemiological studies. I am not an expert on legal matters and my opinion is not well-informed from a legal perspective. However, my opinion is that if most (if not all) of the following can be established in a particular case, then some contribution of the head injury should be acknowledged:
1. The individual was well functioning prior to the head injury, from an occupational and social perspective, i.e. related well with others and functioned well in age-appropriate roles.
2. The head injury produced significant brain trauma, as evidenced by loss of consciousness and/or a period of amnesia and/or cognitive deficits (attributable to the head injury).
3. The first signs of schizophrenia, including social withdrawal and odd behaviours, emerged after the head injury.
4. There is no period of complete normalcy (absence of all psychiatric or neurological symptoms and no cognitive dysfunction) between the head injury and the onset of schizophrenia. (It is possible that there is a return to normalcy in some cases, but the association become more difficult to establish, especially if there is a long interval).
These may be strict criteria, but in the face of controversy, it is better to have strict criteria. They do not deal with the situation in which head injury has arguably worsened psychosis in someone who already has schizophrenia, but this is not the topic of discussion at present. If the above criteria are met, then one should not dismiss the possibility that head injury has made a contribution to the development of schizophrenia, although it should still be considered to be a minor contribution, and so compensated in law.

1. Nielsen AS, Mortensen PB, O’Callagahn E, et al. Is head injury a risk factor for schizophrenia? Schizophrenia Research 2002 55:93-98.
2. Harrison G, Whitley E, Rasmussen F, Lewis G, Dalman C, Gunnell D. Risk of schizophrenia and other non-affective psychosis among individuals exposed to head injury: Case control study. Schizophrenia Research 2006; 88:119-126.
3. Zhang QC, Sahdev PS. Psychotic disorder and traumatic brain injury. Curr Psychiatry Rep 2003;5(3):197-201.
4. Kim E. Does traumatic brain injury predispose individuals to develop schizophrenia? Curr Opin Psychiatry. 2008 May;21(3):286-9.
5. van Winkel R, Stefanis NC, Myin-Germeys I. Psychosocial Stress and Psychosis. A Review of the Neurobiological Mechanisms and the Evidence for Gene-Stress Interaction. Schizophrenia Bulletin 2008; 34 (6): 1095–1105.
6. Howes OD, McDonald C, Cannon M, Arseneault L, Boydell J, Murray RM. Pathways to schizophrenia: the impact of environmental factors. International Journal of Neuropsychopharmacology 2004; 7 (Supplement 1), S7–S13.

7. Tandon R, Keshavan MS, Nasrallah HA. Schizophrenia, "Just the Facts": what we know in 2008 part 1: overview. Schizophr Res. 2008 Mar;100(1-3):4-19. Review.

1 comment:

  1. Interesting post.

    A family history of Schizophrenia or neurodevelopmental issues may perhaps predispose an individuals brain to react this way to TBI,however it is the TBI itself directly that has caused the psychosis symptoms-ie:an organic process.
    Any insult to the temporal lobe areas,whether through disease or injury could definitely result in psychosis or behavioural symptoms.
    A patient having a low level of social/educational functioning pre injury does not automatically mean that they were on the way to psychosis anyway and that the TBI is a "red herring".