The prevention of schizophrenia.doc

the prevention of schizophrenia authors: a. r. yung a; e. killackey b; s. e. hetrick a; a. g. parker a; f. schultze-lutter c; j. klosterkoetter c; r. purcell a; p. d. mcgorry a affiliations: a the department of psychiatry, the university of melbourne, victoria, australia and orygen research centre, parkville, victoria, australiab the department of psychology, the university of melbourne, victoria, australia and orygen research centre, parkville, victoria, australiac department of psychiatry and psychotherapy, university of cologne, cologne, germanydoi: 10.1080/09540260701797803 publication frequency: 6 issues per year published in: international review of psychiatry, volume 19, issue 6 december 2007 , pages 633 - 646 abstract preventive strategies can be divided into universal, selective and indicated prevention and early intervention. universal interventions are directed to the general population. selective approaches are targeted at people who have risk factors for an illness, but who do not show any current signs. indicated approaches target high risk individuals with minimal signs or symptoms foreshadowing mental disorder, but who do not meet diagnostic levels at the current time. early intervention involves treating those with already diagnosable disorder in a timely and optimal manner aiming to decrease the severity of the illness, and reduce secondary morbidity. although universal and selective interventions are not yet viable strategies, indicated prevention and early intervention are now realistic possibilities in schizophrenia. development of methods to identify those at risk of psychosis continues to evolve. promising results in the prevention and delay of transition to psychotic disorder from high risk state have been found. early intervention in schizophrenia, including promotion of early help-seeking, has been shown to reduce the duration of untreated psychosis, which is known to be associated with poor outcome in schizophrenia. early intervention programmes which optimise the care of the first episode have been shown to produce better outcomes than routine management. introduction mental illness is undoubtedly the most important health issue affecting young people (murray & lopez, 1996). mental health and related substance use disorders account for 60-70% of the non-fatal burden of disease among 15-24 year olds (public health group, 2005) and at least 75% of mental disorders among adults commence before 24 years of age (kessler et al., 2005). in the 2000 global burden of disease study it was found that schizophrenia accounted for 2.8% of the years lost to death (ylds) and 1.1% of the disability adjusted life years (dalys) (murray, lopez, mathers, & stein, 2001). the lifetime prevalence of schizophrenia is approximately 1% and the point prevalence is around 0.5%. psychotic illnesses have a peak onset in the late adolescent-early adulthood phase of life (kaplan, sadock, grebb, 1994). left unrecognised, untreated, or poorly treated, psychotic illnesses during this critical developmental period not only lead to considerable personal and family distress and increased severity of illness, but also contribute to poor academic performance, premature exit from school and higher education, unemployment, sustained disability and premature death (mcgorry & yung, 2003; mueser & mcgirk, 2004). the risk of suicide, particularly in early to mid adulthood in those with untreated psychosis and schizophrenia is high (jackson & birchwood, 1996; power, 1999; sane, 2002; who, 2001). a concerted focus on early identification and effective intervention is obviously a public health priority.despite this, preventive endeavours in schizophrenia have long been hampered, in large part, by the pessimistic and fundamentally inaccurate kraepelinian notion of the illness being characterized by progressive and inevitable decline (mcgorry, 1991). until recently, this notion was ingrained in the minds of clinicians (cohen & cohen, 1984), with inevitable iatrogenic effects and an ongoing neglect of recovery-orientated rehabilitation programmes. despite these pessimistic notions, progress has been made over the last two decades towards promoting and adopting a preventive approach to the treatment of schizophrenia and other psychotic illnesses (mcgorry, yung, & phillips, 2002a). this review evaluates the evidence for early intervention and identification of those at risk for developing psychosis with a view towards intervening to prevent or delay development of the illness (yung et al., 1998; 2003).theory and stages of prevention mrazek and haggerty (1994) provide a modern framework for the prevention of mental disorders, describing universal, selective and indicated prevention. universal interventions are directed to a general population group. selective approaches are targeted at people who are at increased risk based on an understanding of the known risk factors and/or aetiology of the disorder, but not showing any current sign of illness. finally, indicated approaches target high risk individuals with minimal but detectable signs or symptoms foreshadowing mental disorder, but who do not meet diagnostic levels at the current time.universal prevention strategies there has been little work investigating the effectiveness of universal prevention strategies in schizophrenia. because they are directed at the whole population, any interventions must be acceptable and without side effects. although universal interventions are defined as those directed to the general population who do not have risk factors, there are some risk factors for mental illnesses in general, and schizophrenia in particular, that are so widespread that their targeting can be considered a universal intervention. for example, growing up in an urban environment has been found to be a risk factor for schizophrenia (van os, hanssen, de graaf, vollebergh, 2002; van os, 2004). the reasons for this are unclear, but explanations include increased exposure to environmental stress such as high traffic density, noise and violence (spauwen, krabbendam, lieb, wittchen, van os, 2006). low social capital including poor social cohesion, exclusion and discrimination may also increase stress (boydell, 2003; van os & mcguffin, 2003) which may be the mechanism by which the risk for psychosis is increased (mortensen et al., 1999; torrey, miller, rawlings, & yolken, 1997). stressful life events in general may increase the risk for psychosis (frangou & murray, 2000). obviously preventing people from living in cities is not a viable preventive strategy, but increasing social capital and decreasing discrimination are possible goals of universal prevention. these strategies are worthy goals in their own right, have no adverse effects and may reduce incidence of a number of mental disorders, not just schizophrenia.other common risk factors for schizophrenia which could also be considered as targets for universal prevention include low socio-economic status, obstetric complications, illicit drug use and childhood trauma. the direction of causality of low socio-economic status and schizophrenia has also not been established. it may be that those who experience psychotic illness drift down through society via unemployment, unstable relationships, social isolation and withdrawal (aro, aro, & keskimaeki, 1995; mcnaught et al., 1997). on the other hand, lower socio-economic status is associated with greater stress, which may increase risk for psychosis (mortensen et al., 1999; torrey et al., 1997). relief of poverty and social marginalisation may therefore be goals of preventive strategies.similarly, use of illicit drugs, particularly cannabis, has been shown to have a definite, albeit modest, causal relationship to onset of psychosis in three large population-based prospective long- term follow-up studies (andrasson, allebeck, engstrm, & rydberg, 1987; arseneault et al., 2002; zammit, allebeck, andreasson, lundberg, & lewis, 2002). reducing population rates of cannabis consumption could be another focus of universal prevention. pregnancy and birth complications are thought to have an impact due to hypoxic or ischaemic neuronal injury (geddes & lawrie, 1995; mcneil, 1995; warner, 2001). childhood trauma has also been associated with an increased risk of psychosis (e.g. greenfield, strakowski, tohen, batson, & kolbrener, 1994), although other prospective studies have found no link with a diagnosis of schizophrenia (e.g. spataro, mullen, burgess, wells, & moss, 2004). reducing these risk factors in the population is therefore potentially of benefit.several of these risk factors are non-specific to psychosis but associated with many psychiatric sequelae. for example, in the case of childhood trauma, there is a demonstrated association between child sexual abuse and a subsequent increase in rates of childhood and adult mental disorders (spataro et al., 2004).thus universal interventions that target these risk factors have broad preventive value for a range of mental disorders, not only psychosis and schizophrenia, making the argument for universal prevention even more compelling. moreover, universal interventions that have intrinsic benefit for the population generally, for example, interventions to reduce the rate of perinatal birth complications (e.g. improved prenatal nutrition, reduction in perinatal incidents) provide further impetus to implement such interventions (mojtabai, malaspina, & susser, in press). it has been argued that a large group of people exposed to a low risk results in a larger incidence rate, than a small group exposed to high risk (rose, 1992; offord, kraemer, kazdin, jensen, & harrington, 1998) such that population based programmes with a large reach ultimately reduce more disorders (brown and liao, 1999, cited in shochet & ham, 2004). they are also less stigmatising and overcome problems of poor recruitment and high refusal during the screening phase often seen in targeted interventions. on the other hand, disadvantages to these programmes include provision of small benefits to individuals, and it is often hard to detect the beneficial effects because of small effect sizes. this is particularly worrying if the major effects are seen in those who are at low risk, while the smallest effects are in those at highest risk. universal interventions are also typically expensive (offord et al., 1998) and difficult to evaluate, especially in the case of low incidence disorders due to sample size considerations (cuijpers, 2003).the rationale for selective and indicated prevention in psychosis empirical evidence suggests that early and effective interventions, rather than universal prevention strategies, are the most cost-effective population health strategies available, and are the focus of the remainder of this review. the onset of psychotic disorders, including schizophrenia, peaks in late adolescence and early adulthood (hafner et al., 1993). it is increasingly important therefore that intervention shifts from universal to selected and indicated prevention, given what is known about the impacts of different social (e.g. alcohol and drug exposures, peer group challenges), genetic (e.g. individual variability in vulnerability to both life stress and drug exposures) and neurobiological (maturation of frontal lobe and limbic structures) risk factors that emerge during the adolescent and early adult periods.the case for selected and indicated prevention in psychosis has been articulated elsewhere (mcglashan, 1996) and increasingly has widespread support (killackey & yung, in press).evidence of effectiveness of selective interventions selective preventive interventions are targeted to individuals or subgroups whose risk of developing disorder is significantly higher than average. risk may be imminent or lifetime (mrzak & haggerty, 1994). thus application of selective prevention requires the identification of risk factors, that is, characteristics which are associated with heightened probability of later onset of disorder (eaton, badawi, & melton, 1995). risk factors is a broad term which is more usefully divided into risk indicators, which denote increased risk but are not causative, and risk modifiers which are associated with causation (tarrant & jones, 1999). for example minor physical anomalies, such as dysmorphic features and skin abnormalities are risk indicators, possibly demonstrating some aspect of abnormal brain development (buckley, 1998; mcneil & cantor-graae, 2000). maternal exposure to influenza (adams, kendell, hare, & munk, 1993) and perinatal brain insults (geddes & lawrie, 1995; mcneil, 1995) are possible risk modifying precursors, that is, they actually influence the degree of risk experienced by the affected individual. alternative but similar terminology has been proposed by kraemer and colleagues (1997), who distinguish different risk factors depending on whether they can be changed or not. they describe risk factors that cannot be changed as fixed markers. these are products or mechanisms that may or may not be causal (such as minor physical anomalies, family history, obstetric complications and delayed milestones). risk factors that can be changed are called variable risk factors, which are then subdivided according to whether their manipulation changes the outcome. if manipulation of the factor changes the outcome it is dubbed a causal risk factor. if changing the factor does not change the outcome, it is called a variable marker. examples of causal risk factors would therefore include illicit drug use and social deprivation (eaton & harrison, 2001), both of which are potentially mutable.childhood behavioural abnormalities, such as over-reactive behaviour in boys and unreactive or withdrawn behaviour in girls (davidson et al., 1999; done, sacker & crow, 1994; done, crow, johnstone, & sacker 1994; reichenberg et al., 2000; 2002) could be considered both risk indicators (manifestations of underlying brain disorder) and risk modifiers or variable risk factors, in that they may be malleable and their treatment may alter subsequent outcome. that is, such behavioural patterns could result in altered environments and experiences for the affected children, which could result in further exposure to risk factors. for example, over-reactive behaviour in an 11-year-old boy may result in him being labelled as a naughty child and may change the behaviour of others towards him. a pattern of conflict with others could then emerge. he may engage in more risk taking behaviours, including substance use. these experiences could result in further abnormalities in the developmental process. thus the abnormal over-reactive behaviour (the risk indicator) becomes the first factor in a series of risk modifiers (yung & mcgorry, 2004).however, despite the recognition of several possible risk factors for schizophrenia, none has a powerful effect and there is an assumption that the disorder is caused by a complex interplay of biological, social and psychological factors. the specificity of most risk factors is low, meaning that most subjects who are exposed to the risk factor do not develop the disorder. these issues make it difficult to demonstrate an effect of selective intervention (cuijpers, 2003).furthermore, risk markers are not sufficient for onset of psychotic disorder. environmental stressors are probably needed to convert underlying vulnerability to manifest disorder, and interaction between liability and stressor is a possible mechanism (spauwen et al., 2006). selected prevention would rely on identifying a group at risk on the basis of risk markers, and targeting them for intervention. however, using these risk markers alone is too narrow a basis for screening and preventive intervention. since individuals would be asymptomatic (at least in relation to psychotic symptoms), albeit displaying a marker of some type, the type of intervention proposed is also not clear cut. thus at this stage, selective interventions are not viable as an effective preventive strategy in schizophrenia.to date, the majority of early intervention work has been undertaken in those with the earliest clinical manifestations of disorder (mrazek & haggerty, 1994). these strategies are most easily categorized as indicated prevention strategies.evidence of effectiveness of indicated interventions the relevance of the prodrome one of the key challenges in attempting indicated prevention in schizophrenia is to determine which signs and symptoms are the precursors to the full threshold disorder. until recently, our knowledge about this was gained from detailed history taking and retrospective studies of patients with schizophrenia (mller & husby, 2000; yung & mcgorry, 1996; 1997). these revealed that a prodromal phase was often evident prior to the onset of frank psychotic symptoms. frequently described were depressed mood, anxiety, irritability and aggressive behaviour, suicidal ideation and attempts, and substance use (yung & mcgorry, 1996), as well as subtle subjective deficits including cognitive, affective and social disturbances. these have been termed basic symptoms (gross & huber, 1996; klosterkotter, ebel, schultze-lutter, & steinmeyer 1996).more proximal to the onset of frank psychotic symptoms, attenuated or subthreshold psychotic symptoms may develop, such as overvalued ideas. perceptual disturbances, such as visual or auditory distortion, can also occur. additionally, deterioration in functioning and behavioural symptoms has long been observed (bowers, 1965; chapman and chapman, 1987; docherty, van kammen, siris, marder, 1978; donlon & blacker, 1973; heinrichs & carpenter, 1985; herz & melville, 1980; meares, 1959).identifying this prodromal phase, in order for intervention to be provided, is the essence of indicated preve