Month: August 2013

Schizophrenia: Does psychological therapy make symptoms worse?

In the field of mental health, schizophrenia is one of the most famous labels out there. It’s often misunderstood, but it’s also relatively common in the general population. Antipsychotics are often criticised for the effects they have on the physical health and mortality of those who take them and emotions on both sides of the argument run rather high.

Psychologists don’t prescribe medication and we often like to think that the treatments we dispense have far fewer side effects. That’s not necessarily the case though – if you’re working with someone who’s traumatised, it’s perfectly possible to retraumatise them through therapy (although we try our absolute best to avoid doing so. But we’re not immune and it would be pompous to pretend that we do nothing but nothing but good.

This paper appearedn recently in The Journal of Nervous and Mental Disease. It was designed to investigate the effect of psychological therapy on symptoms associated with schizophrenia. Schizophrenia symptoms are typically divided into ‘positive’ and ‘negative’ (there is a third category but it simply refers to cognitive functioning). The former are the things we tend to associate with the label: delusions, hallucinations and unusual or disordered thinking. The latter are less flamboyant: they tend to be a certain ‘flatness’ in mood, lack of spontaneous speech and limited pleasure from daily activities. Often, the negative symptoms look very much like depression and they can be the first things you notice.

This study looked the effect of cognitive-behavioural therapy on negative symptoms. It’s an interesting study, because CBT is the recommended intervention fro depression and is one of the recommended interventions for schizophrenia. At the same time, there has been little investigation into the adverse effects of CBT in this population. Psychologists may not be medical doctors, but the old adage of ‘first do no harm’ remains pertinent. CBT was compared to cognitive remediation therapy. For those unfamiliar with either, CBT looks at the patterns between thoughts, feelings and behaviours. It considers how you interpret events and how these interpretations affect the way you respond. CRT is designed to diminish the cognitive decline associated with schizophrenia by providing training in domains such as attention, memory and learning. 198 patients (only 1/3 of those who were eligible) consented to participate and were randomly allocated to receive either CBT or CRT. The effect of random allocation is that you even the sample out – if you knowingly allocate you might skew the groups in terms of age, gender or severity of symptoms. Because we’re all vulnerable to biases, it’s preferable to randomise and to have the randomisation done by machine so that the research team doesn’t know which participants are getting which treatment (this is referred to as ‘blinding’ and the researchers will get the data collected rather than carrying out the treatments as this would obviously mean they knew who was getting what).

So. The results. No one died through suicide during the course of the study, which is good, especially as there is a higher suicide rate associated with schizophrenia. One made an attempt, however, and one had a ‘suicidal crisis’. Those allocated to CBT had slightly elevated scores on a measure of psychotic symptoms compared to those given CRT following treatment, but it was a very light increase; essentially, the risk only increases from a baseline of 1 to 1.1. However, those allocated to CBT were significantly more likely to report elevated depressive (negative) symptoms than their CRT colleagues.

So what do we make of it? Negative symptoms may protect against suicide and some research suggests this is the case. But equally, we don’t want to exacerbate negative symptomatology. In general, CBT had no more ‘side effects’ than CRT: some of those given CBT reported more symptoms at the end of treatment, but their symptoms were no more severe than those of the participants with CRT. Unfortunately, there was no ‘control’ group; a group of people who had similar demographic characteristics to the rest of the sample but who were given no treatment. So although we can state that there is an association between CBT and this increase in symptoms, we can’t attribute one to the other.What this means clinically is that we still don’t know how bad CBT is for your health. What we do know is that it doesn’t prevent those receiving it from going through a crisis (that probably won’t come as a surprise to many) but that, for some people, it may be linked to more negative symptomatology. What clinicians have to do, therefore, is closely monitor the people they see and ensure that crisis and risk plans are both kept updated and implemented as necessary, as well as be prepared to offer more intensive interventions when necessary. Crucially, we need to remember that there are potential risks when engaging in psychological work, just as there are when taking medication.

Female offenders: The role of trauma and mental illness

My professional background (though not my personal, I hasten to add) is in offending and it remains an area which fascinates me. A lot of the research in the area is obviously dedicated to reducing offending, be that in terms of severity or frequency, and a significant amount tries to identify factors which increase the risk of offending, i.e. ‘risk factors’. Recently, an article was published in Psychology of Women Quarterly. The paper investigated the effect of trauma and mental illness in imprisoned women using a mixed methods design. As a general rule, research is divided into ‘quantitative’ (typically employing questionnaires or other measurable outcomes which can then be analysed statistically to ascertain whether there has been a change in whatever has been investigated) and ‘qualitative’, which often uses interviews or focus groups. Crudely, one involves numbers and the other involves words. Both have their pros and cons but it seems to me that quantitative research has more adherents and I sometimes worry that by focusing more on numbers we lose the richness of qualitative research, which can provide extremely in-depth data and which is equally valid, despite the lack of statistics.

This study is relatively unusual in that it combines qualitative and quantitative methods; thus using a ‘mixed methods design’. Up to 90% of women prisoners experience physical or sexual violence in the year preceding imprisonment and many have experienced chronic lifetime abuse. Those abused or neglected in childhood are more likely to offend and to be arrested as adults. For many women, delinquent behaviour can be directly linked to the traumatic event, e.g. running away, using drugs, etc. It’s well-known that trauma experiences are common in those who offend (but, crucially, not all those who experience trauma will go on to offend). Secondly, trauma and mental disorder, be it depression, psychosis or personality disorder, are strongly associated. This paper, then, looks at the effects of both trauma and mental illness (MI) in female inmates.

The authors found that 85% of their sample had a history of substance misuse disorders and that half had a history of post-traumatic stress disorder (PTSD). PTSD can only be diagnosed if you have experienced life-threatening violence or a natural disaster, so rape at knifepoint would warrant a diagnosis but chronic, non-violent sexual assault would usually not. In this case, half the women had had PTSD but even more would have experienced acutely traumatic events without developing PTSD. Half had also experienced a MI such as depression or psychosis, both of which are also associated with trauma experiences. Conclusion: major trauma is very common in female prisoners. Specifically in this sample, 86% had experience sexual violence, and 77% had experienced domestic violence. Only 2 of the sample (of a total of 115) had not been victimised in any way at any point in their lives.

This is all interesting enough, but it’s simply associations; it doesn’t tell you what causes what. Similarly, there’s an association between age and height, but without statistical analysis it’s impossible to see the direction of the relationship. Regression analyses are designed to do just this. They look at factors as predictors and outcomes to show you what the biggest predictor is. They also allow you to control for baseline factors. So for example you could look at the relationship between mental illness and gender (predictors) on reoffending (outcome) but since people who have a history of offending a lot are more likely to reoffend a lot you can might also adjust the analyses for baseline rates of offending to even the analysis out.

Unsurprisingly, women with substance problems were more likely to engage in sex work and driving under the influence of alcohol or drugs. Domestic violence was associated with property crime, drug offending, and commercial sex work and it looks as though these men acted as partners, drug dealers, partners in crime (as it were) and pimps. Witnessing violence was associated with property crimes, engaging in assaultive behaviours and use of weapons, sometimes in self-defence and sometimes in defence of others, such as parents or grandparents. Those who had experiences of caregiver violence were much more likely to run away during adolescence; 4 times as likely as those who were not maltreated. Whilst for some this will allow them to escape being at home, we also know that living on the street is associated with substance use and other offending behaviour, as well as placing the person at greater risk of violence.

This paper looked at a number of mental illnesses but not at anxiety or personality disorders. Up to 70% of inmates have a personality disorder (PD) and we know that trauma is implicated in the development of such disorders. The fact that it wasn’t included means that it’s possible that it s PD which is the underlying factor and not necessarily the trauma experience. Nevertheless, it seems that several life events significantly increase the risk of offending and incarceration, including abusive childhood experiences, violent relationships, criminal environments and self-medicating with substances. It also seems that life events such as running away from home following maltreatment are critical.

So pragmatically, what does this mean? Well, for a start, it seems that MI is common in prison, therefore there need to be more funds allocated to assessing and treating it. Secondly, if people have histories of abuse, neglect and maltreatment, how is prison going to care for them? Will it provide consistent care to try to rebuild people’s abilities to form positive relationships? Given that early experiences are so crucial to adult functioning, how can we identify maltreatment more effectively and intervene to reduce the impact? Finally, how will social care, education, housing and substance misuse services join together to create ‘packages’ of care and support for those at the greatest risk of MI and offending? We try to implement such care packages in the UK (it is of note that this study used an American sample) but we still have a significant female offender population and we know that many of the issues raised in the paper are pertinent on this side of the pond. It has long been known that women offenders have specific needs; it just seems that we’ve still not found a way to attend to those needs as well as we could.

Cocoa: Stopping dementia in its tracks (or possibly not)

This morning I heard of a study which had found that drinking hot chocolate regularly could stop older people from developing dementia. The study, run by a team from Harvard, suggests that cocoa improved blood flow to the brain, resulting in better scores on tasks designed to assess memory. I duly had a look at the paper, published in Neurology.

The study was designed to investigate the effect of cocoa on neural coupling, which refers to the relationship between neural activity and cerebral blood flow. Sixty people were included in the study: half were given two cups of high-flavanol (an antioxidant) cocoa each day and the rest were given low-flavanol cocoa. A number of cognitive assessments were carried out, many of which are also used by clinicians in cases where dementia is suspected. The Mini Mental State Examination (MMSE), for example, assesses cognitive functioning in a range of domains, as does the Trail-Making Test. After 30 days of prescribed cocoa-drinking, the cognitive tests were repeated and a significant improvement was observed on one of them (the Trail-Making Test B). There were no differences reported between the groups, i.e. Trails B scores improved regardless of the quantity of flavanol ingested and no other significant improvements were observed in either group.

Like all studies, this one has its limitations. Whilst it’s a positive that the authors split the sample in half and gave them different quantities of flavanol, it’s usually the case that you have an ‘intervention’ group and a ‘control’ group. In this case, you would have had those who took cocoa for 30 days, perhaps splitting them to take different quantities of flavanol, but you would also have had a similar group – matched for age, gender and health – who took no cocoa. They would be your ‘controls’. That way, you can compare scores across all these individuals at baseline and the end of treatment (in this case end of cocoa) and compare them. The advantage of doing this is that you can see if the improvement was associated with the cocoa or if it was simply a spontaneous improvement which was also observed in the control group.

In this study, there was no control group, so it’s difficult to presume that the observed improvement was due to taking cocoa. In any case, an improvement on one test of cognitive function may be interesting, but it won’t necessarily have any real-world impact. The Trail-Making Test assesses a range of functions, but the test is the sort of thing we use clinically, it’s not something you would have much cause to do in daily life. It’s a bit like the Brain Training games which teach you to remember long lists of words – your skills in that domain may improve, but realistically that’s only useful if you want to be able to recite your shopping list verbatim at any moment. I’d be more excited if the improvement had been seen on the MMSE since that is a much more global test of functioning. As it is, this was a small study which didn’t have a control condition and that seriously limits the extent to which we can draw conclusions from the findings. If you’re considering stocking up on the Green and Blacks simply to stave off dementia, I’d think again.

In which I set this blog only moderately high standards

Mental health has long been a topic shrouded in mystery, known only to those who have experience of it, whether as professionals or service users (or, indeed, as both). I’ve always thought that those of us who do have experience of the area have an obligation to remove that shroud; to make it more accessible and less ethereal.

Over the past few years, I’ve noticed that the portrayal of mental health and, in particular, the public dissemination of research into mental illness, has often been problematic. Research ‘findings’ are published in newspapers and on the radio but most of us have little understanding of the scientific method and this limits our ability to digest and critique that research. In my view, anything that contributes to the general public’s understanding of science is good. Unfortunately, I haven’t got the qualifications (or the time) to debunk all the dodgy claims that are made, but I do know a thing or two about mental health, and I’m no stranger to conducting research. This, then, is my humble attempt to contribute to the public understanding of psychology and psychiatry. I hope to discuss stories and research presented in the media, and perhaps some of my own musings too. Debate is most welcome and I look forward to your contributions.