To the Editor: Dr. Mendlowicz and collaborators raise an important issue regarding the difficulties in detecting the psychopathological boundaries of mood and cognitive and anxiety disorders, particularly in schizophrenic subjects. These lifetime comorbidities were not described in detail in our article. Of note, we found no significant differences between schizophrenic subjects with (N=29) and without (N=51) comorbid social anxiety disorder in the lifetime prevalence of comorbid mood disorders (24.1% versus 21.5%). Current major depressive disorder was an exclusionary diagnosis at enrollment.

Although the Structured Clinical Interview for DSM-IV is appropriate for the diagnosis of depression in schizophrenia, the Hamilton Depression Rating Scale may not be optimal for assessing depression severity in this population. Although the Liebowitz Social Anxiety Scale seems to give a consistent measure of the severity of social anxiety disorder in both schizophrenic as well as anxious patients, the Calgary Scale may be preferable over the Hamilton depression scale in evaluating depression in schizophrenia.

The Liebowitz Social Anxiety Scale describes which and how severe the social anxiety-derived behaviors are, but it does not help us to understand how these behaviors have developed (i.e., as the effect of impaired social cognition or social affect). Also, the Liebowitz Social Anxiety Scale does not describe the subjective expression of these symptoms (i.e., rigidity, trembling, language difficulties, blushing, sweating). This rating scale evaluates anxiety and social pervasiveness dimensionally, not the subjective experience that represents the substrate of the socially anxious behavior.

We argue that a "fully dimensional" framework might be considered. From a "syndromic" point of observation, social anxiety disorder in nonschizophrenic patients may be close to depression, considering the comorbidity rates. In schizophrenic patients, the social anxiety experience resembles more closely the psychopathological constellation associated with the so-called "self-centrality" (i.e., nondelusional self-referential attitude), as described in the subjective experience "basic symptoms" model (1) of schizophrenia and subjective unawareness. Particularly, impairment in language competence in some schizophrenic patients might represent the basis of the "social incompetence" that drives social anxiety.

We refer to the original concept of "endophenotypes" (2) as internal phenotypes discoverable by a "biochemical test or microscopic examination." The method available for endophenotypic analysis has advanced, and our current armamentarium includes neurophysiological, biochemical, endocrinological, cognitive, and neuropsychological (including configured self-report data) measures. Reports of subjective experiences may turn out to be potentially useful as a psychopathological organizer of schizophrenia spectrum disorders. Subjective experiences represent a strictly neuropathological expression, substrate-close disorders from which clinical symptoms and behaviors arise. Psychopathological emphasis on schizophrenia-associated disorders, such as social anxiety, may also help to integrate the search for neurodevelopmental mechanisms in schizophrenia with developmental psychology research on the ontogenesis of the self.

In this search for newer models, we need to incorporate the impact of treatments. Our group is now investigating the pharmacogenetics of treatment response to clozapine in schizophrenic patients with and without comorbid social anxiety disorder and would be glad to establish collaborations with other researchers in this area.