IT has been suggested that one can “predict” schizophrenia at an individual level.
Features posited as being part of these so-called ultra-high risk (UHR) criteria include risk factors (eg, a family history of schizophrenia), symptoms (eg, attenuated positive psychotic symptoms) and longitudinal trajectory (decline in functioning). Added to this, mostly, is the requirement that the individual be “help-seeking”, though this is a loose term and raises questions about pathways to care.
Early work using such criteria suggested an impressive “transition” rate to a full-blown psychotic disorder of around 40%. However, most “transitioned” very soon after ascertainment and came from “enriched” clinical groups.
A recent meta-analysis reported a 1-year transition rate of 22%, but more recent clinical trials (see below) suggest a lower rate.
Also, this needs to be seen in the context that around 10%–15% of people will at some stage in their lives experience psychosis-like symptoms, but this does not mean that they have or will develop a disorder, or that they need treatment.
Another twist is that a transition rate of 16% was found in a clinical sample of young people who didn’t meet any UHR criteria.
Even if one could reliably ascertain which individuals have an ultra-high risk of transitioning, are there interventions that can reduce the likelihood of transition?
I am aware of eight published studies that have attempted such an intervention using a controlled experimental design. Six of these were recently summarised. An early study of medication and cognitive behaviour therapy (CBT) (n = 59) favoured the intervention at 6 months (10% transitioned v 36% in the control group), but not at 12 and 48 months; a study of CBT (n = 58) showed reduced transition rates for the intervention group at 12 months (6% v 22%–30%); a trial of olanzapine versus placebo (n = 60) showed no significant difference between interventions, as did a trial of CBT v supportive psychotherapy (n = 51).
A remarkable study (n = 81) showed a dramatic effect of treatment with polyunsaturated fatty acids (4.0% v 27.5% at 12 months), but, to my knowledge, this has not been replicated.
Finally, a Melbourne study of 115 subjects at ultra-high risk of psychosis showed no significant differences in 6-month transition rates between CBT plus antipsychotic medication, CBT plus placebo, and supportive psychotherapy plus placebo (4.7% v 9.1% v 7.1%).
The two more recent studies deserve special mention. In one, 128 individuals referred to as being in an “early initial prodromal state” were randomly assigned to a comprehensive “integrated psychological intervention” or supportive counselling. Transition rates favoured the intervention arm at 12 months (3.2% v 16.9%) and 24 months (6.3% v 20.0%).
However, the assessment tool differed from most comparable studies; the intervention was multifaceted and intensive; therapist and group contact time between the intervention and control arms was not controlled for; and the low proportion of actual participants selected for randomisation from the pool of potential participants suggests a highly non-generalisable sample.
Most recently, 288 individuals meeting UHR criteria were randomly assigned to cognitive therapy and mental state monitoring versus monitoring alone. Only 23 participants transitioned (8%) and there was no difference in rates between the two groups (odds ratio, 0.73; 95% CI, 0.32–1.68).
This brief overview suggests a convergence of two issues: first, the predictive value of current UHR criteria is relatively poor. Second, the recent, adequately powered and methodologically robust intervention studies show modest or no effect of assorted interventions (pharmacological, psychological or a combination thereof) on transition rates.
There is also the very real problem that more harm than good might be done with such interventions, including labelling, stigma and exposure to potentially toxic medications. Even the shunning of labels such as “schizophrenia” cannot detract from the stigma regrettably still associated with mental illness, although provision of services in general health and low-stigma environments is welcome.
It is very well established that many antipsychotics carry a major burden in terms of metabolic problems and may have direct neurotoxic effects, and their adverse effects have been noted in studies of individuals meeting UHR criteria.
Although the attempt to reduce transition rates might seem rather tenuous, the psychological interventions investigated do seem to have the potential to help people “cope” with their psychosis-like experiences.
Taking a leaf from the UHR proponents’ book, perhaps we should be focusing more energy on “psychological strengthening” of at-risk individuals, rather than primarily trying to reduce transition rates.
Unless, of course, fish oil is truly the answer, in which case the dolphins in Hitchhikers’ guide to the galaxy got it just right!
Professor David Castle is professor of psychiatry at St Vincent’s Hospital, Melbourne.
Posted 21 May 2012