17 February 2012. A new study published online January 2 in the Archives of General Psychiatry, led by Mark Weiser of Sheba Medical Center in Tel Hashomer, Israel, finds that, in the general population, self-reported attenuated psychotic symptoms (APSs) confer risk for later psychosis, but aren’t useful to predict who will subsequently convert to a full-blown psychotic illness.
A diagnosis of schizophrenia or a related psychotic disorder is often preceded by a lengthy prodrome, during which the patient exhibits milder, though still distressing, symptoms. Although these attenuated symptoms are thought to confer a high risk for clinical psychotic illness, not all patients with these symptoms eventually develop a major psychotic disorder. There is a strong link between APSs (such as suspicious beliefs, magical thinking, or unusual perceptual experiences) and later development of psychotic illness in the literature (Poulton et al., 2000; Dominguez et al., 2011); however, these studies are limited by the use of non-clinician interviewer evaluations for diagnosis. In the current study, Weiser’s group utilized a more reliable and more conservative outcome measure—subsequent psychiatric hospitalization—to assess the link between APSs and psychotic illness.
Weiser’s study bears attention from the research community, according to Barbara Cornblatt, director of the Recognition and Prevention Program of Zucker Hillside Hospital in Glen Oak, New York. Cornblatt, who was not associated with the study, told SRF that its significance lies in its sample: “a population resource that is totally unique and can provide information you can’t get anyplace else.”
First author Nomi Werbeloff and colleagues utilized data from an epidemiological study performed in Israel in the 1980s using 4,914 individuals aged 25 to 34 years at the time of baseline assessment. All subjects with a diagnosis of psychotic illness at baseline were subsequently excluded. APSs were self-reported using the false beliefs and perceptions subscale of the Psychiatric Epidemiology Research Interview, with approximately 15 percent of the sample population reporting at least one strong APS within the last year, and over half (57 percent) reporting at least one weak APS within the same timeframe.
The researchers then merged the epidemiological data set with a list of all psychiatric hospitalizations in the country as of 2007, using the National Psychiatric Case Registry, to assess whether the APSs were associated with later risk of hospitalization for nonaffective psychosis (including schizophrenia and schizoaffective disorder, but excluding bipolar disorder). Subjects were followed for an average of 24 years.
Increased risk, but no predictive power
Werbeloff and colleagues report that a higher average APS score was associated with a greater risk of later hospitalization for a nonaffective psychotic disorder, with an odds ratio of 3.6 (indicating that for every unit increase in mean APS score, a greater than threefold increase in risk of hospitalization occurred). With each additional weak and strong APS reported, the risk for later nonaffective psychosis increased by 30 percent and 41 percent, respectively. Additionally, similar findings were observed using hospitalization for pure schizophrenia as the outcome measure, with an odds ratio of 4.1.
Self-reported APSs also increased the risk of hospitalization for other non-psychotic psychiatric illnesses, though to a lesser extent than for nonaffective psychosis. A Kaplan-Meier survival curve demonstrated that, following strong APSs, nearly all hospitalizations for nonaffective psychosis occurred within a much narrower time window following baseline than those for a non-psychotic psychiatric illness (five years vs. 12 years, respectively), which suggests that, to some degree, APSs may be specific for nonaffective psychoses.
Although over half the subjects reported some form of APS at baseline, only 0.5 percent were later hospitalized for a nonaffective psychotic illness, consistent with previous reports using other outcome measures (e.g., van Os et al., 2000; Kendler et al., 1996). Thus, not surprisingly, APSs are substantially more common in the general population than psychotic disorders, but unfortunately can’t be used reliably to forecast subsequent psychiatric illness. The 12-month conversion rate among high-risk individuals, who often have a family history of psychosis and decline in cognitive function in addition to experiencing APSs, is around 10 percent (Cannon et al., 2008; Ruhrmann et al., 2010). The substantially lower conversion rate in the general population observed by Werbeloff and colleagues suggests that variables such as APSs that are used to predict conversion in high-risk clinics may not be relevant for predicting psychosis in the general population.
Interestingly, poor social functioning and anxiety disorders at baseline emerged as variables that significantly interacted with APSs in increasing the risk of hospitalization for nonaffective psychosis, consistent with recent data implicating these variables in risk for psychotic illness (Cornblatt et al., 2011; Freeman, 2007). Although the current study was not able to determine the nature of these interactions (e.g., does poor social functioning increase the conversion from APSs to psychosis, or does it increase the rate of APSs?), future studies may shed light on the role of these variables in transition to psychotic disorders.
One potential limitation of this study, as noted by the authors, is the relatively advanced age of the population sampled. The onset of psychosis typically occurs during late adolescence or early adulthood—well before the average baseline age of 29 years in this sample. Given that subjects who converted to psychosis prior to the baseline assessment were excluded, an earlier baseline time point would provide a more complete picture of the relationship between APSs and psychotic disorders. An additional factor that requires clarification in future studies concerns the risk of APSs for subsequent affective psychoses, such as bipolar disorder and depression with psychosis, that were not examined in the present study.
To classify or not to classify?
Identification of patients at high risk for developing psychotic illness has been the focus of a substantial amount of research (see SRF related news story), as well as controversy. The current study may inform a question that is currently on the minds of clinicians and researchers alike: should the new Diagnostic and Statistical Manual of Mental Disorders (DSM-5), slated for publication in 2013, include an "attenuated psychosis syndrome" diagnostic class (Carpenter and Van Os, 2011)? This very question was the subject of a 2009 SRF Online Discussion).
On one side of the debate are those who argue that a separate diagnostic class is warranted since it would serve a patient population that is, by definition, distressed and, in many cases, help seeking. A new diagnostic category, proponents say, would facilitate early treatment, thereby improving outcome. On the other hand, opponents contend that low rates of conversion to psychotic illness, risk of stigma resulting from diagnosis, and the potential administration of unnecessary medication outweigh the benefits of a separate diagnostic class.
So, how does the current study inform this debate? According to Cornblatt, the data support both sides of the argument: “This paper illustrates the difficulties inherent in adapting risk factors from a selected, enriched, high-risk population for use with the general clinical populations [that the DSM is intended for]," she said. "At the same time, it suggests that APS symptoms are meaningful in increasing the risk for nonaffective psychosis.”—Allison A. Curley.
Werbeloff N, Drukker M, Dohrenwend BP, Levav I, Yoffe R, van Os J, Davidson M, Weiser M. Self-reported Attenuated Psychotic Symptoms as Forerunners of Severe Mental Disorders Later in Life. Arch Gen Psychiatry. 2012 Feb 10. Abstract