4 May 2009. Researchers who gathered in San Diego 27-28 March for the International Congress-Emory University Cognition Satellite Meeting spent an afternoon probing the lifetime course of schizophrenia. The session, chaired by Philip Harvey of the Emory University School of Medicine in Atlanta, Georgia, focused on whether cognition worsens or remains stable during the earliest stages of illness, and in late life. In Part 1 of this story, we report on the talks earlier that day on episodic memory, attentional control, and translational issues.
The first speaker, Larry Seidman, of Harvard Medical School in Boston, Massachusetts, reported that, even before the first psychotic episode, subjects who later developed schizophrenia show large neuropsychiatric deficits that may appear as early as age four. On average, they have an IQ of about 95 before they become ill. He and his colleagues found that almost all of the effect of schizophrenia on IQ is present at the first episode.
Seidman and colleagues used a case-control sample from the National Collaborative Perinatal Project, a large, prospective longitudinal study, to explore premorbid intellectual functioning in schizophrenia. They found modest differences on some cognitive domains, particularly in attention-working memory and verbal-conceptual ability, between those who later developed schizophrenia and those who did not. Subjects who eventually developed affective psychoses showed milder deficits than those who developed schizophrenia. Seidman suggested that prodromal neurocognitive deficits might help identify people who need treatment.
According to Michael Davidson of the University of Tel Aviv in Israel, findings based on a military psychiatric hospital registry indicated that most prodromal patients receive no mental health care for their symptoms until they are hospitalized with a psychotic episode. In other research, he found that those who would later develop schizophrenia performed worse than controls on cognitive tests; they became impaired at least one year before their stay in a psychiatric hospital. However, using this information to predict who will develop schizophrenia is complicated by the presence of similar deficits in most mental disorders. Furthermore, about 15 percent of the general population has neurocognitive scores similar to those seen in schizophrenia, and about 2 percent regularly have psychotic experiences.
Barbara Cornblatt, from Zucker Hillside Hospital in Glen Oaks, New York, presented findings from the Recognition and Prevention Program, a treatment and research program for 12- to 22-year-olds at clinical high risk for schizophrenia and bipolar disorder. These youngsters show symptoms suggestive of the illness, but at levels insufficient to warrant a diagnosis. She found that prodromal subjects who later developed psychosis showed greater cognitive deficits than did normal controls, and their deficits remained unchanged even as the disease emerged. High-risk subjects who did not become psychotic showed intermediate deficits. Social function and role function—two domains largely independent of cognition—also stayed constant during the prodrome. Cornblatt believes that the deficits stem from a genetically based biological vulnerability and require intervention regardless of whether the at-risk youths become ill.
Ty Cannon, of the University of California at Los Angeles, described changes in brain structure and function that occur during the prodromal period and the early stages of psychosis. During the teen years, pruning of extra synapses occurs as part of normal development, but Cannon suggested that overzealous pruning occurs in the prefrontal and temporal regions in first-episode patients with schizophrenia and during the prodrome. In fact, he and his colleagues found that lower gray matter density in the prefrontal cortex predicts conversion to psychosis. Furthermore, their diffusion tensor imaging work reveals white matter changes in young adults with recent-onset schizophrenia that correlate with impaired working memory. Prodromal subjects also show these changes, but to a lesser extent.
In other work from Cannon and colleagues, first-episode patients who did well on working memory tasks showed greater activity in the dorsolateral prefrontal cortex than controls. This suggests that first-episode patients may have needed to work harder to achieve the same outcome as controls. Similarly, Cannon’s ongoing research hints that prodromal patients perform worse than healthy controls on verbal working memory tasks. In particular, those who go on to develop schizophrenia show the most activity in the dorsolateral prefrontal cortex.
Untangling aging, practice, and disease
Turning to adults who already have the disease, Eric Granholm, of the University of California, San Diego, said that only a few studies have examined the stability of cognitive deficits from middle age to late life in schizophrenia, and they have yielded inconsistent findings. Usually, studies assume that subjects with schizophrenia will benefit from practice to the same extent as others, but if aging, on top of existing disease-related deficits, impairs patients’ ability to improve with practice, clinical trials may produce misleading results.
Granholm’s study of subjects 45 to 75 years old, including those with and without schizophrenia, teased apart the effects of practice and aging by conducting assessments at randomized intervals and using mixed statistical models. On neuropsychiatric tests, older subjects with schizophrenia benefited from practice less than younger patients and healthy controls, particularly on working memory tasks. The study reported no such effect on visual information processing tasks.
The last speaker, Philip Harvey, spoke about the course of cognition and disability in older patients with schizophrenia. Unlike earlier longitudinal studies that focused on younger patients, Harvey’s study tested patients who had a mean age of 60. Subjects had stayed in an institution for varying lengths of time, although all lived in the community at the time they completed a battery of cognitive and functional tests. With practice, those with a short institutional stay improved their performance on tests of cognition and daily living skills, whereas longer-stay patients did worse over time. Harvey cautioned that if the patients had not been split into groups, the results would have shown no change over time.
Harvey’s second study examined the sensitivity of neuropsychiatric tests to changes with age in schizophrenia. He pointed out that neuropsychiatric test norms assume normal age-related decline. In a sample of 50- to 85-year-olds, many neuropsychiatric tests showed no interactions between age and a diagnosis of schizophrenia. However, both the Digit Span Distraction Test and the Trail Making Test, Part A, found age-related changes in performance in subjects with schizophrenia, but not healthy controls. According to Harvey, the findings suggest that traditional neuropsychiatric tests may not detect age-related declines that occur in people with schizophrenia.—Victoria L. Wilcox.