Study by Roman Kotov, PhD and Colleagues Calls into Question the Diagnosis of Schizoaffective Disorder

October 24, 2013 – A study by Roman Kotov, PhD and colleagues published in JAMA Psychiatry called into question the validity of the diagnosis of schizoaffective disorder, which occupies a place between schizophrenia and mood disorders in current diagnostic systems. They proposed instead a diagnostic scheme with a single boundary between schizophrenia and mood disorders, as originally put forward by Emil Kraepelin 114 years ago. They suggested that people displaying symptoms of both schizophrenia and mood disorder would best be described as having schizophrenia with a comorbid mood disorder rather than given the label of schizoaffective disorder.

They arrived at their conclusion by applying advanced statistical techniques to data obtained from 526 participants in the Suffolk County Mental Health Project. Using nonlinear statistical models to evaluate the associations between initial symptom course and global outcome measures obtained 10 years after participants entered the study, the researchers found that the relations between symptoms of psychosis or depression and outcome measures were linear, but they observed a sharp dichotomy in outcome measures related to the nonaffective psychosis ratio, which reflects relative duration of psychosis outside of mood episodes to the total duration of the illness.

This ratio is a key component in the diagnosis of schizoaffective disorder. According to the DSM-5, in psychotic mood disorders, nearly all psychosis has to occur within the context of a mood episode (the ratio is essentially zero). In schizoaffective disorder, both nonaffective psychosis and mood episodes are prominent, whereas in schizophrenia, nonaffective psychosis predominates. In other words, this ratio separates schizoaffective disorder from mood disorders on one end and schizophrenia on the other.

If schizoaffective disorder falls on a continuous spectrum between schizophrenia and mood disorder, the association between the nonaffective psychosis ratio and outcome measures would be linear; if schizoaffective disorder represents a construct distinct from both schizophrenia and mood disorder, the association would likely be characterized by two breaks, one marking it off from schizophrenia and the other from mood disorders. Dr. Kotov and his colleagues found, however, only one distinct nonlinear discontinuity indicating a single dichotomy between schizophrenia and mood disorder.

In a companion editorial, Kenneth Kendler, MD of the Medical College of Virginia and a member of the DSM-5 work group on mood disorders, commended Dr. Kotov and his colleagues for demonstrating that the idea of a smooth continuum from psychotic mood disorders to schizophrenia can be rejected, but withheld judgment on their conclusion that schizoaffective disorder is not a valid diagnostic category. He suggested instead a “Scottish verdict” of not proven. This is consistent with the suggestion by the authors of the original article that their conclusions require verification in other samples and with a variety of validators.

The article, titled Boundaries of Schizoaffective Disorder: Revisiting Kraepelin was published online October 2, 2013 in JAMA Psychiatry (formerly Archives of General Psychiatry). It adds to a substantial literature evaluating the validity of schizoaffective disorder—and as in most studies, it did not find evidence for a distinct disorder—but it used novel statistical techniques that test boundaries with full rigor. Laura Fochtmann, MD; Eduardo Constantino, MD; Gabrielle Carlson, MD; and Evelyn Bromet, PhD were among the article’s co-authors.