Diagnosis and Treatment of Mood Disorders That Co-occur With Schizophrenia

Although descriptions of mental illness have been made throughout recorded history, rigorous attempts to define patient-experienced symptoms, as opposed to observable signs, were not made in Western medicine until the 19th century (6). Psychiatrists at that time were embedded in an intellectual system that described the functions of the mind on the basis of the "pure reasoning" of philosophers such as Kant (7) and Locke (8), whose descriptions of the mind focused on the "senses," mental "representations," and the concepts or belief systems that were built from those representations in conjunction with "innate" logical functions. This emphasis led to a fairly straightforward development of clinical terms for dysfunctions of the "thought" process, or problems of perceptual malfunction or irrationality—for example, delusion, hallucination, and dementia.

A more difficult challenge for psychiatrists in the 1800s was the creation of concepts such as "mood" and "affect." Western medicine and philosophy had long conceptualized a separate aspect of the human being, also recognized among nonhuman animals, usually defined as "passions" (9,10) or "humors" (11), which operated independently of reason and were described primarily in terms of volition and energy level. Indeed, excitement of the passions was recognized to sometimes lead to insanity—or irrationality—and therapeutic manipulation of the passions was one early form of treatment for disorders of the mind (12).

However, conceptualization of the passions in terms of processes that themselves could have pathologies was possible only after philosophies were constructed that considered affective function as a "primary, autonomous, and irreducible faculty of the mind" (6,13). Armed with the conceptualization of affect as a discrete faculty of the mind, and encouraged by the 19th-century Romantic movement, with its emphasis on subjective experience, psychiatrists developed a terminology of affective symptoms, including concepts such as mood and depression. The concepts of mania and melancholia—previously defined by particular delusions, hallucinations, and activity levels—were revised to emphasize the affective symptoms of elation, expansiveness, and sadness as key components of these syndromes.

Nineteenth-century psychiatrists soon found that there were many possible combinations of psychotic disorders—disorders marked by symptoms such as delusions, hallucinations, and disintegration of linear thought—and affective disorder—defined by the evolving concept of affective syndromes such as depression and mania. Discussions soon developed on the means of determining when mania or depression might be secondary or primary to psychotic disorders (14,15) and whether mood disorders with comorbid psychosis should be considered as separate from mood disorders without psychosis (16).

Of special interest to our current concepts of mood disorders in the context of schizophrenia—especially with regard to the idea of "negative" symptoms—was the eventual conceptualization of volition as an independent mental faculty. In 1863 Kahlbaum (17) described volition as a third mental faculty and defined a category of psychiatric disturbances that could be described as dysfunctional conditions of volition—separate from disturbances of the intellect and the emotions.

In the midst of a widening appreciation of the complexity of these systems and their interactions, Kraepelin (18) introduced a major division of the functional psychoses based on whether the patient had a cyclical or a chronic course of illness and a good or a poor prognosis and whether precipitants played a role in the onset of symptoms. This division gained predominance in psychiatry, no doubt in large part because of Kraepelin's emphasis on recognizing putative disease processes rather than having to focus on symptom clusters and syndromes, whose etiologies were complex, confusing, and ultimately unknown. Thus the stage was set for 20th-century psychiatry and the development of the concepts of schizophrenia—Kraepelin's "dementia praecox"—and bipolar disorder—Kraepelin's "manic-depressive illness"—as two separate disease processes.

Bleuler (14) and Schneider (19) took Kraepelin's basic notion of a chronic psychotic disease, which Bleuler renamed as schizophrenia, and developed descriptions of particular symptoms or primary traits—that is, symptoms that were always present—that they believed were pathognomonic of the "disease." Curiously, despite the strong pull to separate thought disorders and mood disorders in 20th-century psychiatry, Bleuler nevertheless recognized disorders of affect—in particular, the deterioration of affect—as one of the three primary or fundamental symptoms of schizophrenia. Bleuler also recognized that other types of mood disorder, such as depression and mania, are often present among persons who have schizophrenia. However, he considered depression and mania as two of many possible secondary, or "accessory," syndromes of schizophrenia.

Concurrently with Bleuler's and Schneider's attempts to define "the schizophrenias," English clinicians defined particular symptom clusters and syndromal variants that could be used to define the manic-depressive illnesses, including nonpsychotic versions of these illnesses (20). In 1933 a third diagnostic entity called schizoaffective psychosis was introduced by Kasanin (21) to describe psychiatric illness marked by prominent affective and psychotic features that appear in a cyclical course.

The 1970s saw the development of tentative, or "experimental," specific diagnostic categories by Feighner and associates (22) and Spitzer and associates (23). These categories were intended to provide criteria to be used by clinicians who wanted to define major depressive, manic, and schizoaffective syndromes as well as the "diseases" of schizophrenia, schizoaffective disorder, and bipolar disorder. These authors derived their criteria by asking a number of colleagues what they considered to be the essential symptoms or signs of these syndromes or illnesses. The component symptoms and signs that constituted the syndromes or illnesses were derived from the theories of Kraepelin, Bleuler, Schneider, and the English school of Gillespie and colleagues (20), but the concepts for making differential diagnoses were derived principally from the work of Feighner and associates (22), who emphasized course of illness in contrast with the cross-sectional diagnoses that were popular in the United States in the 1960s and early 1970s.

Psychiatric illnesses that presented with comorbid psychotic and affective features were classified according to whether the psychosis occurred only during the peak of affective episodes—bipolar disorder and major depressive disorder—or whether the affective disorders appeared only after psychotic episodes—schizophrenia with postpsychotic depression (24). The diagnosis of schizoaffective disorder was more formally defined in the late 1970s and early 1980s as a type of schizophrenic condition in which comorbid mood disorders were present most of the time (25,26).

The American Psychiatric Association, in successive versions of the Diagnostic and Statistical Manual of Mental Disorders, adopted the research criteria developed by Feighner and Spitzer and further clarified the notions of syndromes, episodes, and disorders. The nomenclature that has developed allows identical clusters of symptoms to be considered as an episode—for example, major depressive episode and manic episode—or a syndrome—for example, major depressive syndrome and manic syndrome—on the basis of whether the cluster of symptoms occurs in the context of bipolar disorder or major depressive disorder or in the context of schizophrenia or schizoaffective disorder, respectively.

Furthermore, DSM-IV has eliminated any criteria related to "pathognomonic" psychotic symptoms with regard to bipolar disorder and schizoaffective disorder, now placing all emphasis on the time course of the affective syndromes or episodes in relation to the psychotic features of the illness. In a seeming return to Bleuler's tolerance for mood disorders in the diagnostic category of schizophrenia, DSM-IV also allows depressive syndromes and manic syndromes to co-occur during the course of schizophrenia, as long as the total duration of these syndromes does not add up to a significant portion of the total duration of the psychotic illness, in which case the diagnosis of schizoaffective disorder is given.

The ontologic weakness of our current categorical system is that the affective syndromes that lie at the heart of this system were arrived at by consensus polling and were not based on rigorous attention to either biological or neurophysiologic substrates that might help to more accurately define disease processes in psychiatry. Schizophrenia, on the other hand, has seen some modifications in its definition, at least in DSM-IV, partly because of studies that, through factor analyses of many possible symptoms, have shown three primary dimensions in schizophrenia, characterized by positive, disorganized, and negative symptoms (27).

Some researchers have suggested that other dimensional models of schizophrenia exist besides the three-factor model at the heart of the DSM-IV diagnosis (28). For example, the work of Kay and Sevy (29), which was based on an assessment of 30 diverse pathologic symptoms in a sample of patients who were diagnosed as having schizophrenia according to DSM-III, strongly supports a four-dimensional model of schizophrenia composed of a positive dimension, defined by grandiosity, delusions, and unusual thought content; a negative dimension, defined by emotional withdrawal, blunted affect, passive or apathetic social withdrawal, poor rapport, poor attention, and lack of spontaneity and flow of conversation; a depressive dimension, consisting of sad mood, alterations in sleep and appetite, and feelings of guilt; and an excited dimension, consisting of symptoms of excitement, tension, and poor impulse control (29).

If one were to superimpose a patient's manic syndrome, as defined by DSM-III or DSM-IV, onto the dimensions proposed by Kay and Sevy, one would immediately notice that individual symptoms of mania actually correspond to entirely different dimensions—grandiosity fits into the positive syndrome, poor attention into the negative dimension, changes in sleeping patterns into the depressive dimension, and excitement and poor impulse control into the excited dimension.

In a parallel type of study, Kendler and colleagues (4) applied a latent class analysis to 343 patients who had been treated for psychotic and affective disorders and, using a 21-item rating scale, determined six underlying disease syndromes. These syndromes were validated on the basis of distinct demographic, familial, clinical, and course data that were not used in the latent class analysis. Their study used modern computer and statistical methods to define a nosology in much the same way that Kraepelin had once used empirical observation and grouping of patients to derive his binary psychosis model. The six types of illness proposed by Kendler and colleagues suggest that the spectrum of psychotic and affective disorders is more complex than Kraepelin suggested. The types of illness proposed by Kendler and colleagues are classic schizophrenia, bipolar schizomania, major depression, schizodepression, schizophreniform disorder, and hebephrenia.

Although this classification system delineated six different disease processes, it is of great interest that some syndromes defined in DSM-IV, such as mania and depression, were strongly represented in more than one of the disorders. This multirepresentation suggests that our currently defined mood syndromes, mania and depression, may actually each be observable in several pathologic processes. For example, elevated mood, excessive activity, reckless acts, and pressured speech were all prominent symptoms noted by Kendler and colleagues in a substantial portion of patients who had bipolar schizomania, hebephrenia, or schizophreniform disorder.

Finally, as neuroscience has developed, more detailed hypotheses have emerged to describe the neural substrate from which all psychiatric disorders ultimately derive. Anatomical studies, functional imaging studies, and more detailed cognitive studies have begun to piece together the neural mechanisms of emotion (3). As described by Derryberry and Tucker (3), psychiatry's historical tendency to search for biological systems that underlie a given syndrome, such as depression or mania, ultimately fail to appreciate the "interdependence of the multiple layers of the neural hierarchy." The hippocampal system, for example, connects putative behavioral inhibitory and excitatory systems with other neural mechanisms in the paralimbic cortex and neocortex, which process complex mental representations of the self and the social environment. Medications provide yet another useful tool for separating distinct pathologic processes that, from an observational standpoint, might appear identical.

In summary, our current nosologic system, DSM-IV, although useful for separating different groups of patients who have different time courses of mood syndromes and psychotic syndromes, is not likely to accurately reflect the underlying pathology and disease structure of many patients whom we currently define as having schizophrenia or schizoaffective disorder. Depression and mania, the predominant mood syndromes that we diagnose in the context of schizophrenia, may actually represent composites of different dimensional systems (29) rather than discrete pathologic entities. Finally, there may be different disease processes that are not defined accurately by DSM-IV and that are lumped together in our DSM-IV categories for schizophrenia and schizoaffective disorder, which individually may produce syndromes such as mania and depression.

The development of the research diagnostic criteria (23) and the series of iterations that led to DSM-IV have provided psychiatrists with the ability to reliably define certain phenotypes—groups of patients who have characteristic syndromes and time courses—and, subsequently, to test the efficacy of particular treatments for those clinical entities. The criteria of Feighner and colleagues (22) and Spitzer and colleagues (23), on which DSM-IV is based, were intended to suggest experimental definitions of syndromes and illnesses. As newer models or definitions of syndromes and illnesses are introduced to the field (4,30,31), psychiatry may move beyond the DSM-IV classifications to diagnostic descriptions that will have more validity. Nevertheless, our current literature of evidence-based treatments of mood disorders that co-occur with schizophrenia is relevant only in the context of our current diagnostic system. Psychiatrists who want to base their treatments on what has been proven in the literature must therefore first master the rules of diagnosing according to DSM-IV.

Proper therapeutic choices for the treatment of depression and mania among patients who have schizophrenia begin with a proper differential diagnosis. Given that substance abuse and some physical health problems, such as epilepsy, often co-occur with schizophrenia, a thorough examination to rule out a possible role of substance abuse and organic causes of depressive and manic syndromes should be a standard component of the assessment and treatment of all patients with schizophrenia. In the case of depression, one must first determine whether some of the symptoms are secondary to negative symptoms or medication effects—for example, oversedation, the direct effect of dopamine blockade, and extrapyramidal symptoms. Once the cause has been determined, the symptoms can be treated appropriately.

Treatments for these causes of secondary symptoms of depression might include an increase in the dosage of the neuroleptic that is being administered or a switch to an atypical neuroleptic agent, if the symptoms are really the negative symptoms of schizophrenia; a decrease in the dosage of the neuroleptic agent or a switch to a less sedating neuroleptic, if the primary cause of some of the depressive symptoms is oversedation; administration of anticholinergic medications; or a switch to a neuroleptic agent that has fewer side effects. Careful treatment of these aspects of schizophrenia will often reduce the number of depressive symptoms, allowing the physician to consider a wider range of interventions to treat whatever depressive symptoms remain.

When negative symptoms and side effects of medications are optimally treated, depression in a patient with schizophrenia might be modified to only dysthymic mood, which might be amenable to supportive individual or group psychotherapy (48,49,50,51). It may then be possible to avoid adding additional psychotropic medications, such as antidepressants, to the patient's regimen.

The same principles apply to the treatment of manic syndromes among patients who have schizophrenia. The clinician should first determine whether a few or several of the manic symptoms—for example, psychomotor agitation, irritable or excited affect, and grandiose delusions—might actually be due to undertreatment of the core psychotic illness, in which case increasing the dosage of the neuroleptic agent or switching to a more effective neuroleptic agent would be the first move. The clinician should assess whether some of the symptoms are in fact side effects of current medications—for example, akathisia appearing as psychomotor agitation or insomnia, in which case the side effects should be treated by decreasing the dosage, switching to a neuroleptic agent with fewer side effects, or administering adjunctive medication to treat the side effects of the medication that is responsible for the akathisia. As with depression, optimization of the neuroleptic medication and treatment of side effects may alleviate several of the "manic" features, eliminating the need to treat a full manic syndrome.

Once other causes of depressive and manic symptoms have been ruled out or addressed therapeutically, several options remain for treating depression or mania if that is still deemed necessary. Supportive psychotherapy, either individual or group based, should be implemented to allow close monitoring of the affective symptoms, to help the patient recognize these symptoms, and to help the patient make adjustments that will position him or her to overcome the affective aspect of the illness. For example, therapy for depression can aim to provide an empathic, supportive environment in which the patient can mourn the losses incurred by the illness and in which more realistic goals can be set. For patients who are experiencing manic syndromes, therapy can be aimed at identifying and avoiding risk factors that induce or worsen the syndrome, increasing the patient's insight into his or her capacity to become manic, highlighting the problems that manic syndromes can cause, and emphasizing the need to address the onset of mania with a psychiatrist.

A substantial body of literature exists on pharmacotherapeutic options for the treatment of depression that co-occurs with schizophrenia. Neuroleptic agents have been shown to have a clear effect on decreasing depression ratings among persons who have schizophrenia in addition to depressive syndrome, and these agents may be the treatment of choice when the patient is actively psychotic (52,53).

The evidence is conflicting regarding older theories that antidepressants worsen the course of schizophrenia during the actively psychotic phase, presumably through stimulation of the catecholamine system. For example, Kramer and colleagues (52) showed that combining antidepressant medications with neuroleptics for the treatment of actively psychotic patients with both schizophrenia and depression did not alleviate the depression and actually exacerbated some of the positive symptoms of schizophrenia. However, Müller-Siecheneder and associates (54) showed that a combination of haloperidol and amitriptyline was associated with marked reductions in both psychotic and depressive symptoms among persons with schizophrenia who had both disorders.

On the other hand, there is no controversy about the value of antidepressant adjunctive medication for patients with schizophrenia who experience a full depressive syndrome after their psychosis has remitted. Siris and colleagues (55,56) conducted randomized, double-blind, controlled studies showing that in the treatment of postpsychotic depression among patients with schizophrenia who were receiving a standing dosage of neuroleptic medication, a daily dose of 200 mg of imipramine was significantly more efficacious than placebo in ameliorating depression and in preventing relapse of depression. Moreover, the combination of imipramine with the neuroleptic medication had a significant effect in preventing relapse of psychosis among these patients. Other antidepressants as well as lithium have shown some promise for use in the treatment of depressive symptoms in the course of schizophrenia (2,57,58). However, these studies were conducted in small samples and had mixed results.

Atypical antipsychotics—in particular clozapine, olanzapine, and risperidone—which have direct effects on the serotonergic system in addition to specific dopaminergic receptors, have proved to be especially efficacious in the treatment of depressive symptoms that co-occur with schizophrenia (39). Clozapine has shown specific efficacy in reducing both suicidality and the number of suicide attempts among persons with schizophrenia (59). Olanzapine has been shown to reduce the one-year rate of suicide attempts compared with haloperidol in the treatment of patients with schizophrenia (60).

In contrast with the number of studies of the treatment of depression in the context of schizophrenia, studies on the pharmacologic treatment of manic syndromes, or even manic symptoms, among these patients are lacking. In a literature review by Levinson and colleagues (2), no controlled studies on pharmacologic treatment of manic symptoms or syndromes were found. Only one report was cited as being relevant—a study of the benefits of adjunctive lithium treatment among persons with schizophrenia who were also taking neuroleptic medication. In that report, some of the patients with schizophrenia who had symptoms that might be described as hypomanic showed a favorable response to treatment with a neuroleptic agent plus lithium.

Keck and associates (39), although they do not focus on manic syndromes, point out that risperidone has been shown in a case-control study to be more efficacious than haloperidol or placebo in reducing "hostility" ratings among patients with schizophrenia. Another study showed that both olanzapine and haloperidol were more efficacious than placebo in reducing hostility scores on the Brief Psychiatric Rating Scale (61). The paucity of studies on the treatment of manic symptoms in persons who have schizophrenia may be due to the tendency of many in psychiatry to overlook the fact that according to DSM-IV, patients with schizophrenia may have manic syndromes during the course of their illness. Many such patients are automatically classified as having bipolar-type schizoaffective disorder.

Given that the line between schizoaffective disorder and schizophrenia with comorbid mood syndrome is so difficult to draw, it is also worth noting the treatments for schizoaffective disorders, both bipolar and depressed types, that have been studied. Patients with schizoaffective disorder of either type have not, for the most part, shown any improvement in treatment response when lithium has been added to standard neuroleptic treatment (2). Curiously, no studies with clear criteria for diagnosis of schizoaffective disorder or with control designs have been conducted to test the efficacy of anticonvulsants, such as carbamazepine and valproate, in the treatment of either type of schizoaffective disorder. However, the results of some very small open-label studies have suggested that divalproex may be useful as an adjunct to neuroleptic treatment (62).

Finally, several studies have been conducted of the efficacy of atypical antipsychotics for the treatment of schizoaffective disorder. The results of these studies suggest that clozapine may be even more efficacious in the treatment of schizoaffective disorders, especially the bipolar type, than it is in the treatment of schizophrenia (63,64). Risperidone has shown mixed results for treating schizoaffective disorders, although it appears to have greater efficacy for the depressed type than for the bipolar type (65).

Depressive and manic symptoms occur ubiquitously among persons who have schizophrenia, often as full-blown syndromes. Factor analysis and latent class analysis suggest that for at least some patients with schizophrenia, these syndromes are actually core factors of the psychotic illness. The results of modern neuroanatomical and pharmacologic studies strongly suggest that the neural substrates that underlie mood disorders and thought disorders are not as distinct as the philosophical models of the 18th century and the diagnostic categories that derived from those models in the 19th and 20th centuries had hypothesized.

Although mood syndromes have major relevance for the outcomes of patients with schizophrenia, especially with regard to factors such as suicidality, course of illness, and hostility, our field in general has not aggressively pursued studies to determine the optimal treatments of depression and mania that co-occur with schizophrenia. Mania in the context of schizophrenia, although a valid diagnostic concept according to our current diagnostic system, DSM-IV, is especially underrecognized and understudied, perhaps because of the way the diagnostic criteria for schizophrenia and bipolar disorder were established historically.

Diagnosis of mood disorders that co-occur with schizophrenia requires a careful separation of confounding factors, such as exacerbations of the primary psychotic illness itself, side effects of medications, substance abuse, and organic factors. When mood syndromes are definitely present, tricyclic antidepressants are beneficial for treatment of depression and prevention of relapse of psychosis once the active psychosis has been treated. Lithium and anticonvulsants may be of use in treating manic syndromes, but no well-designed studies have been conducted to test their usefulness scientifically.

Atypical antipsychotics have been clearly shown to reduce depressive symptoms, including suicidality, and to reduce hostility, although their effect on treatment of full depressed and manic syndromes has not been measured among patients with schizophrenia. Atypical antipsychotics are also efficacious in the treatment of schizoaffective disorders. Clozapine especially shows promise for the treatment of bipolar-type schizoaffective disorder. More research on the efficacy of mood stabilizers other than lithium as well as antidepressants other than tricyclics in the treatment of mood syndromes that co-occur with schizophrenia and schizoaffective disorder is very much needed.

This work was supported in part by grant MH-01453 from the National Institutes of Health.

Dr. Escamilla is with the neurogenetics laboratory of the department of psychiatry at the University of Texas Health Science Center, 7703 Floyd Curl Drive, San Antonio, Texas 78229-3900 (e-mail, [email protected]).