Late-Onset Schizophrenia-Like Psychosis

To the Editor: At the risk of being isolationist, I want to take issue with the international consensus on late- and very-late-onset “schizophrenia-like psychosis” (1). I see no heuristic advantage and potentially great therapeutic harm in labeling patients with such illnesses schizophrenic, no matter the prefix or suffix. Conceptually, I thought the international consensus was that schizophrenia is likely a developmental disorder resulting from an interaction between a genetic predisposition and adverse gestational, labor and delivery, or perhaps neonatal effects on the developing nervous system. Furthermore, this initial lesion expresses itself in childhood with neuromotor, cognitive, and emotional deficits and in late adolescence and young adulthood as a nonaffective psychosis, most likely with negative symptoms and a variable but generally poor long-term course (2). How are we to reconcile this concept with the schizophrenia that develops after 40 or 60 years of age? If we cannot, why call the late-onset psychosis “schizophrenia”? What is wrong with Kraepelin’s term “paraphrenia”?

In the recent article on this topic, the consensus group summarized information about these late-onset psychoses: they are more likely to occur in women and to be associated with mood disturbances and with positive rather than negative features (particularly visual hallucinations in elderly subjects) and less likely to be familial for schizophrenia and more likely to be familial for mood disorders. Subjects with these disorders are also more likely to commit suicide. This pattern does not seem to fit the standard set 31 years ago by Robins and Guze (3) for establishing the diagnostic validity of disorders of unknown etiology. Why isn’t the late-onset illness a form of mood disorder and the very-late-onset illness part of a heterogeneous group of illnesses that includes delirium, dementia, depression, and the like?

More important, what useful clinical purpose does it serve to call these patients “schizophrenic”? Too many patients are already needlessly exposed to antipsychotics and their risks. Words affect thinking. A patient labeled “schizophrenic” is more likely to receive an antipsychotic and less likely to receive an antidepressant, a mood stabilizer, or ECT than a patient labeled as having an atypical mood disorder or paraphrenia. There used to be an international consensus that illness results from an imbalance in four body humors. Consensus without conceptual logic or clinical utility is meaningless and dangerous.