Psychosis Nonconverters Present Clinical, Research Dilemma

Individuals identified as at risk for psychosis but who did not convert to psychosis within two and a half years showed improvement in symptoms and functioning from baseline, but continued to perform significantly more poorly compared with normal controls.

At-risk adolescents and young adults enrolled in the North American Prodrome Longitudinal Study 1 (NAPLS 1) who did not convert to psychosis within 30 months ("nonconverters") showed significant improvement in attenuated positive symptoms and negative symptoms as measured by the Scale of Prodromal Symptoms and in social and role functioning as measured by the Global Assessment of Functioning.

The results were presented by Jean Addington, Ph.D., one of the principal investigators of NAPLS at the University of Calgary, in an adddress at the International Congress on Schizophrenia Research (ICOSR) in April in Colorado Springs, Colo.

NAPLS 1 is an eight-site study directed by the National Institute of Mental Health that has followed more than 300 individuals classified as "high risk" according to the Structured Interview for Prodromal Symptoms (SIPS). Criteria for prodromal psychosis include a family history of psychosis accompanied by deterioration in functioning and/or the presence of attenuated positive symptoms of psychosis; the latter include unusual thought content, delusional ideas, suspiciousness or persecutory ideas, grandiose ideas, perceptual abnormalities or hallucinations, and disorganized communication.

Addington reviewed published data on rates of conversion to psychosis in prevention studies, but also presented new information on 76 subjects who had met criteria for entry into the NAPLS study, had never taken antipsychotic medication, and had not converted to psychosis by two years.

The improvement over baseline in symptoms and functioning among these nonconverters largely occurred in the first year of follow-up, Addington said. "Over a two-year period, from the first year to the second year of follow-up, there was some slight improvement in scores, but most of the really significant improvement occurred in the first year," she said

But in a comparison with 111 healthy controls, Addington reported that the nonconverters continued to perform significantly more poorly. "So even though there was an improvement in symptoms and functioning, they are still not back to normal levels," she said.

The subject of psychosis risk has become a major focus among schizophrenia researchers, a fact reflected in the number of presentations, symposia, and workshops on the subject at this year's congress and in the existence of a satellite International Prodromal Research Network that meets concurrently with the congress.

Several large-scale studies around the world have yielded data on conversion rates of individuals identified as at risk for psychosis. In NAPLS, 35 percent of the more than 300 subjects who met criteria for prodromal psychosis had converted to a full-blown psychosis within two years—data that were originally reported in the January 2008 Archives of General Psychiatry.

That rate was matched by the Prevention Through Risk Identification, Management, and Education (PRIME) study, which also reported a 35 percent conversion rate at 12 months.

As debate has grown around the possible inclusion of a diagnostic category in DSM-5 for psychosis risk, along with concern voiced in some quarters about "false positives" and possible stigmatization of adolescents, researchers have turned their attention to those individuals who do not convert to psychosis. Better insight into the demographics, baseline characteristics, and longitudinal course of individuals who originally meet at-risk criteria but who do not convert to active psychosis may help advance understanding of the prospective course of at-risk individuals and improve predictive rates.

Conversion Rates Falling

Intriguingly, Addington and others at the congress reported that conversion rates among at-risk samples appear to be falling somewhat. When research on the prodrome began, most studies were replicating the 35 percent conversion rates found in NAPLS, but some other studies are now showing conversion rates of 15 percent to 27 percent, she said.

Why this is happening is unclear; it may represent the success of early detection and prevention efforts that involve various forms of treatment, or it may represent a "dilution effect" from the inclusion in samples of a growing number of individuals with less-severe symptoms.

Addington said that among nonconverters in the NAPLS sample, just 5.4 percent still met the criteria for prodromal psychosis at two years, but that 31 percent (40.5) had at least one positive symptom at the attenuated level. "So what appears to be happening is that some people were still experiencing symptoms, but had a decrease in symptoms so that they no longer met criteria," she said. "In fact, nobody had positive symptoms at any point in the follow-up greater than the symptoms they had at baseline."

Thirty-two percent of the nonconverters had an anxiety disorder at two years, and 14 percent had depression. Twenty-nine percent had a range of Axis II disorders.

Treatment protocols across the eight sites in NAPLS 1 varied and may have included case management and close monitoring of symptoms or treatment with antidepressant or antipsychotic medications. (However, the data Addington presented at the ICOSR on nonconverters excluded any who had ever had an antipsychotic.)

A fuller description of NAPLS 1 methods and protocols appeared online in the January 25, 2007, Schizophrenia Bulletin.

New Category for Psychosis Risk?

The ongoing debate about inclusion of a category in the DSM-5 for psychosis risk was an undercurrent throughout the congress and emerged in a question-and-answer period following Addington's presentation.

She concluded her remarks by saying that the data on nonconverters—and the fact that conversion rates appear to be falling in prevention studies—challenge researchers to think about issues of diagnostic boundaries and the specificity of measures used to categorize individuals as "at risk." And she expressed skepticism about the inclusion of a diagnostic category for psychosis risk in DSM-5.

In remarks to Psychiatric News after the congress, Addington said that better predictive criteria are needed, as is a better understanding of the clinical course of people who remain below the threshold of active psychosis. She added that NAPLS 2, a sequel to the first study now enrolling 360 at-risk subjects and 180 normal controls, will have more rigorous data about both converters and nonconverters.

But at ICOSR, Addington was challenged after her presentation by William Carpenter, M.D., chair of the psychosis work group for DSM-5. He noted that the issue of a category in DSM-5 for psychosis risk is very controversial and that opinions within the work group itself are not uniform (see ‘Attenuated Psychosis Syndrome’ Subject of AJP Editorial).

But in response to Addington's concern about the danger of creating a classification for people, some of whom will not convert to psychosis, Carpenter pointed out that the nonconverters in NAPLS 1 were among a help-seeking population that continued to require psychiatric treatment, as evidenced by the persistence of Axis I and II diagnoses. "Where is the harm?" he asked.

In the May American Journal of Psychiatry, Carpenter coauthored an editorial exploring all of the aspects of the "at-risk" controversy with Jim van Os, M.D., a leading researcher who has opposed the inclusion of a psychosis risk category.

Also responding to Addington at the congress, Thomas McGlashan, M.D., a pioneer in research on psychosis risk and prevention, said that nonconverters continue to require the close monitoring of mental health professionals.

"What these young people need is to be watched very closely," he said. "They need a clinician to be with them, and that requires an investment of time and effort on the part of the clinician. And in order to be paid for that time and effort, you need to have a diagnosis."

"Prediction of Psychosis in Youth at High Clinical Risk: A Multi-Site Longitudinal Study in North America" is posted at http://archpsyc.ama-assn.org/cgi/content/full/65/1/28.