62: Disorders of Thought and Volition: Schizophrenia

• Diagnosis of Schizophrenia Is Based on Standardized Clinical Criteria

  • The Symptoms of Schizophrenia Can Be Grouped into Positive, Negative, and Cognitive

  • Schizophrenia Is Characterized by Psychotic Episodes

• Both Genetic and Nongenetic Risk Factors Contribute to Schizophrenia

• Neuroanatomic Abnormalities May Be a Causative Factor in Schizophrenia

  • Loss of Gray Matter in the Cerebral Cortex Appears to Result from Loss of Synaptic Contacts Rather Than Loss of Cells

  • Abnormalities in Brain Development During Adolescence May Contribute to Schizophrenia

• Antipsychotic Drugs Act on Dopaminergic Systems in the Brain

• An Overall View

The success of neurobiology in providing insights into perception, cognition, and more recently emotion has inspired increasingly sophisticated biological investigations into disorders of thought and mood. In this chapter and the next we examine the four most serious disorders of thinking and mood: schizophrenia, depression, mania, and the anxiety states. These disorders involve disturbances in thought, self-awareness, perception, affect, volition, and social interaction.

In addition to being scientifically challenging, mental illness such as schizophrenia is of great social importance. Tragically this illness results in lifelong disability. The World Health Organization counts schizophrenia as one of the most significant contributors to disease burden (defined as healthy years of life lost to illness) worldwide. Fully 5% of people with schizophrenia commit suicide. Many more are homeless. The vast majority are unable to function successfully in school or in the workplace. Before the advent of psychopharmacologic therapies, schizophrenia and the mood disorders accounted for more than half of all hospital admissions in the United States. Even now schizophrenia accounts for approximately 30% of all hospitalizations.

The pattern of symptoms of schizophrenia are remarkably similar in all countries and cultures. The average prevalence worldwide ranges between 0.5 and 1%; the male-female ratio is 1.4:1. Diagnosis is usually made during late adolescence or early adulthood with the emergence of full symptoms, but in retrospect the illness begins far earlier with prodromal symptoms.

In medicine the understanding of a disease, and therefore its diagnosis, is ultimately based on identification of (1) etiological factors (such as microbes, toxins, or genetic risks) and (2) pathogenesis (mechanisms by which etiologic agents produce disease). Unfortunately, the etiology and pathogenesis of most mental disorders have not been determined. As a result, psychiatric diagnoses still rely on the patient's description of symptoms, the examiner's observations, a detailed natural history (the course of the illness over time), and the response to treatment.

This approach to psychiatric diagnosis began at the turn of the 20th century with the work of Emil Kraepelin in Germany. Influenced by Rudolf Virchow, the German pioneer of cellular pathology, and by Thomas Sydenham, the English clinician who focused attention on the natural history of medical diseases, Kraepelin studied mental disorders as specific disease processes. Even without knowledge about the etiology and pathogenesis, of diseases affecting thought, emotion, and behavior, he argued, such diseases could still be distinguished on the basis of signs, symptoms, and natural history.

Of course, the presentation of a single sign or symptom is not in itself evidence for disease because it may occur in healthy people. But when certain signs and symptoms occur together they form a syndrome, a condition that can be distinguished from normal behavior or from other clusters of signs and symptoms. The natural history of a disease is studied by tracing the onset of signs and symptoms in patients' lives and how they change with time. Thus a syndrome can emerge at a characteristic age or it can follow a characteristic clinical course. For example, Kraepelin recognized that most patients with schizophrenia (which he called dementia praecox) do not recover the level of functioning they had prior to the onset of the disease, whereas most patients with mood disorders experience cycles of relapse and at least partial recovery.

Since the 1980s the diagnosis of psychiatric disorders has been based on standardized criteria that have made diagnosis more reliable. Two different clinicians applying standardized criteria for schizophrenia are very likely to arrive at the same diagnosis. Nevertheless, without etiological or pathophysiological data and lacking objective tests, current diagnostic systems such as the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) of the American Psychiatric Association cannot define disease states in scientifically verifiable terms. With progress in such areas as genetics and neuroimaging, it eventually should be possible to arrive at objectively verifiable and thus valid diagnostic criteria for mental disorders.

The Symptoms of Schizophrenia Can Be Grouped into Positive, Negative, and Cognitive

It is useful to subdivide the symptoms of schizophrenia into three clusters because each may reflect different aspects of the pathophysiology and because each responds differently to the medications presently used. Positive or psychotic symptoms include mental phenomena that do not occur in healthy people, such as hallucinations and delusions. Negative (or "deficit") symptoms result from impairment of normal functions and can include blunted emotional responses, withdrawal from social interactions, impoverished content of thought and speech (Box 62–1), and a lack of motivation.

A third symptom cluster includes cognitive abnormalities—sometimes described as "disorganization symptoms." These symptoms impair working memory and executive functions—the ability to organize one's life. These cognitive symptoms typically persist even during otherwise successful treatment with medication and are thought to be significant contributors to long-term disability. Interestingly, cognitive symptoms can be found to some degree in persons at very high risk of developing schizophrenia but who have not yet experienced hallucinations or delusions, and in otherwise healthy relatives of patients with schizophrenia, suggesting that cognitive symptoms reflect genetic predispositions to schizophrenia.

The medications used to treat schizophrenia are called antipsychotic drugs and are most effective at diminishing the positive symptoms as well as psychotic symptoms that occur in mood disorders. None of the medications reliably benefits the cognitive symptoms.

Schizophrenia Is Characterized by Psychotic Episodes

The most dramatic manifestations of schizophrenia are psychotic symptoms, including hallucinations and delusions. Hallucinations are percepts that occur in the absence of appropriate sensory stimuli and can occur in any sensory modality. In schizophrenia the most common hallucinations are auditory. Typically, a patient hears voices, but noises or music are also common. Sometimes the voices will carry on a dialog and frequently are experienced as bullying and derogatory. Occasionally, voices will issue commands to the patient that can create a high risk of harm, including suicide. Neuroimaging studies of subjects experiencing auditory hallucinations suggest that areas normally involved in the processing of language are recruited during hallucinations. These include Broca's area in the frontal lobe and Wernicke's area in the superior temporal lobe of the cerebral cortex (see Chapter 60).

Delusions are firm beliefs that are not realistic and not explained by the patient's culture. They can be so powerful that sufferers cannot (or refuse to) compare their beliefs to what is actually happening in the world. Delusions can be quite varied in form. For some patients reality is distorted: The world is full of hidden signs meant only for them (delusions of reference), or they are being closely watched or persecuted (paranoid delusions). Others experience bizarre delusions, for example that some entity is inserting or extracting thoughts from their brain or that their dental fillings are radio transmitters broadcasting what they say to nefarious groups. Psychotic symptoms can also occur in mood disorders and drug-induced delirium, but the other symptoms and clinical course of those states are not consistent with schizophrenia.

The full emergence of schizophrenia is often preceded by a period of early symptoms. In this prodromal period the patient can behave eccentrically, become socially isolated, exhibit blunted affect, poverty of speech, a poor attention span, and lack of motivation. Once the disease is fully manifest, periods of florid psychosis typically occur, accompanied by markedly disordered thinking and abnormalities in the regulation of emotion. These periods of overt psychosis are interspersed with periods of residual symptoms. After the first few episodes the patient rarely returns to full normal functioning.

Schizophrenia, like many other mental illnesses, runs in families. As early as 1930 Franz Kalman in Germany studied familial patterns of transmission and concluded that genes contribute significantly to schizophrenia. Three major strategies have been used to quantify the contribution of heredity to the risk of schizophrenia and to understand how genetic risk is transmitted.

In one strategy the rate of concordance for schizophrenia in monozygotic twin pairs, whose DNA sequences are 100% identical, is compared with that in dizygotic twin pairs, whose DNA sequences are on average 50% identical. Assuming that the familial environment is roughly identical for both types of twin pairs, then if genes play a significant role the concordance rates should be higher among monozygotic pairs than among dizygotic pairs. In fact, monozygotic twins have a concordance rate of nearly 50% for schizophrenia, whereas dizygotic twins have a concordance rate of approximately 15% (slightly higher than that for ordinary siblings, which also have on average 50% genetic identity).

Although these rates suggest an important role for genes in schizophrenia, they also demonstrate that genes are not completely determinative. If they were, the concordance rate for monozygotic twins would be 100% as it is in Huntington disease for example. Thus factors other than inherited DNA sequence, such as new mutations, epigenetic modification of DNA, environmental factors, and stochastic factors occurring during brain development, play a role in converting inherited genetic vulnerability into the disease.

To separate genetic factors and environmental influences more clearly, Seymour Kety, David Rosenthal, and Paul Wender examined children who were adopted at or shortly after birth in Denmark, a country where very accurate family and health records are kept. They found that the rate of schizophrenia in the biological family of an adoptee was much more strongly predictive of schizophrenia than the rate in the adoptive family. Kety and his colleagues also observed that some of the blood relatives of schizophrenic adoptees exhibited some symptoms of schizophrenia, such as social isolation, suspiciousness, eccentric beliefs, and magical thinking, even though they did not have full-blown schizophrenia. These symptoms are part of what is now called schizotypal personality disorder. Kety and his colleagues did not possess modern understandings of working memory, but it would have been interesting to know whether their sample also exhibited the cognitive abnormalities now documented in some relatives of people with schizophrenia. Overall the schizotypal symptoms are thought to be a mild, nonpsychotic form of the disease.

More recently, unaffected monozygotic twins and even siblings of patients with schizophrenia have been found to exhibit some neuroanatomic abnormalities similar to those with the disease. In one magnetic resonance imaging (MRI) study monozygotic twins discordant for schizophrenia had similar deficits in the dorsolateral prefrontal cortex and superior temporal gyrus.

Studies by Irving Gottesman of extended pedigrees of Danish patients with schizophrenia also support the importance of genes. Gottesman noted the correlations between the risk of schizophrenia in relatives and the percentage of the total genetic material each relative shared with the patient. He found a greater lifetime risk of schizophrenia among first-degree relatives (parents, siblings, and children, who share 50% of the relatives' DNA sequences) than among second-degree relatives (aunts, uncles, nieces, nephews, and grandchildren), who share 25% of their DNA sequences with the patient. Even third-degree relatives (who share only 12.5% of the patient's DNA sequences) were at higher risk for schizophrenia than the 1% of the population at risk for this disease (Figure 62–1).

With the advent of modern genomic technologies during the last decade, progress has been made in identifying variations in DNA sequence that contribute to the risk of schizophrenia. As with many common disorders, risk for schizophrenia has proven to be genetically heterogeneous, with no single gene proving necessary or sufficient. Two forms of genetic variation have been associated with schizophrenia: Variations in single nucleotide bases, and larger chromosomal deletions, duplications, or translocations. In most cases schizophrenia appears to result from the action of a large number of genes together with environmental risk factors. In a minority of cases schizophrenia risk is markedly elevated by chromosomal abnormalities such as a microdeletion on chromosome 22q11.2.

As with any genetically influenced disease, understanding how certain sequences at particular loci in the genome confer risk should provide important clues to pathophysiology, and thus treatment development. Because schizophrenia may result from abnormal developmental processes in the brain, knowing the time during development and adulthood when the genes that predispose to schizophrenia are expressed in the brain would be valuable, as it would suggest different avenues for investigation and might also suggest optimal times for therapeutic intervention.

The complexity of schizophrenia genetics is illustrated by a well-studied chromosomal translocation that was discovered in a large Scottish family. This translocation between chromosomes 1 and 11 inactivates a gene that came to be called Disrupted in Schizophrenia-1 (Disc-1) that appears to have significant roles in brain development. Within this multigenerational family, individuals who inherited the translocation exhibit serious mental illness but not necessarily schizophrenia. Some have bipolar disorder and others have major depression. Thus truncations in Disc-1 must interact with other genes and nongenetic factors to determine the ultimate phenotype.

Attempts to identify objectively measurable components of schizophrenia and other psychiatric disorders have focused on what have been called intermediate phenotypes or endophenotypes. Intermediate phenotypes may represent measurable structural brain abnormalities—cognitive abnormalities (such as deficits in working memory) measured by their neural correlates on functional neuroimaging—or they may represent measurable neurochemical abnormalities. If intermediate phenotypes can successfully be identified, they may simplify the search for risk genes because they may help identify more homogeneous populations for study than those identified by clinical symptoms and interviewing alone.

The search for modifiable environmental risk factors has also proved daunting because some environmental factors that correlate with the disease may be a result rather than a cause of schizophrenia and others may be proxies for the actual, but as yet undiscovered risks. For example, a consistent relationship has been found between schizophrenia and low socioeconomic status. However, the evidence suggests that schizophrenia itself impairs occupational and social success, leading to downward socioeconomic drift, rather than the alternative notion that stressors associated with poverty contribute to the disease. Other environmental risk factors, including season of birth, urban birth, maternal exposure to viral illness, paternal age, and perinatal complications, have been identified in population studies. Understanding the aspects of urban birth that might contribute to the risk of schizophrenia poses a significant challenge.

Schizophrenia is characterized by certain abnormalities in brain anatomy that can be seen with structural and functional magnetic resonance imaging (fMRI). Thinning of specific areas of the prefrontal, temporal, and parietal cerebral cortex has been observed in many studies (Figure 62–2). The thinning of the prefrontal cortex is most pronounced in the dorsolateral prefrontal cortex, the brain region most critical for working memory.

Thinning in the temporal lobe has been traced to a loss of gray matter in the superior temporal gyrus, the temporal pole, the amygdala (amygdala reductions may be limited to males), and the hippocampus. These regions are normally involved in integrating cognition and emotion. The loss of gray matter is counterbalanced by an increase in the volume of the cerebral ventricles (Figure 62–3).

Structural abnormalities in the brain, such as loss of cortical gray matter, have been correlated with functional abnormalities both in cognitive performance tests and studies with positron emission therapy (PET) or fMRI. Impairment of functions that are dependent on the prefrontal cortex have been particularly well documented. For example, patients with schizophrenia have deficits in working memory and cognitive control, which are correlated in functional neuroimaging studies with lack of activity in the dorsolateral prefrontal cortex.

Loss of Gray Matter in the Cerebral Cortex Appears to Result from Loss of Synaptic Contacts Rather Than Loss of Cells

The observed loss of volume in the frontal and temporal cortical regions is not the result of cell death (loss of cell bodies) but rather a reduction in dendritic, axonal, and synaptic processes (neuropil). As a consequence, the density of cells in the cerebral cortex increases. More cells per unit volume and less total gray matter contribute to enlargement of the ventricular spaces.

Like the prefrontal and temporal cortex, the thalamus also appears smaller in patients with schizophrenia compared to nonaffected individuals. But cell counts in postmortem tissue suggest that, unlike the cerebral cortex, there may be loss of cell bodies in the mediodorsal nucleus of the thalamus. Because cells of the mediodorsal nucleus send their axons to the dorsolateral prefrontal cortex, loss of these axonal terminals could in turn contribute to the reduction of cortical dendrites and the dendritic spines that usually receive these thalamocortical connections.

Pyramidal neurons, the most common type of excitatory neuron in the neocortex, receive excitatory input from the thalamus on dendritic spines. Thus the reduction in dendrites and dendritic spines (Figure 62–4) would likely signify a loss of synaptic contacts in the dorsolateral prefrontal cortex in schizophrenia. The loss of synaptic connections in this region could possibly explain the impairment of working memory and executive function that characterizes schizophrenia (Figure 62–5).

Abnormalities in Brain Development During Adolescence May Contribute to Schizophrenia

Because schizophrenia first occurs typically in late teenage years or in the early twenties, symptoms may be triggered by abnormalities in late stages of brain development. Early adulthood is an important period of brain development, as the brain matures in response to a variety of influences. These range from gonadal steroids in adolescence to stressful life experiences such as separating from parents and siblings for college or military service, or becoming independent by taking on adult responsibilities such as employment and sexual relationships.

During this period critical life events are accompanied by synaptic pruning that is part of the selective maintenance of those synaptic connections that are used effectively during normal brain development. Synaptic pruning may be particularly important in the prefrontal cortex. Moreover, the pruning coincides with major changes in dopaminergic neurotransmission in this brain area during late adolescence. The timing of these processes is consistent with the implication of both prefrontal cortex and the dopaminergic system in the pathogenesis of schizophrenia (Figure 62–5).

Although slow continued loss of gray matter in prefrontal and temporal cortex has been observed after diagnosis, cortical abnormalities and ventricular enlargement are generally observed at the time of first diagnosis, suggesting that the pathogenic processes underlying schizophrenia have been active long before psychotic symptoms emerge.

Comparisons of middle-aged and older monozygotic twins that are discordant for schizophrenia (and in which the unaffected twin serves as a control) have shown that the severity of gray matter deficits in the prefrontal cortex correlate with severity of symptoms, not with the duration of illness. Of course, such studies cannot tell us whether the deficits were fully present at the onset of symptoms. To address this issue Judith Rapoport conducted a longitudinal study of those rare individuals with onset of schizophrenia in childhood and documented a correlation between progression of gray matter deficits and ventricular enlargement with time and duration of illness. In these subjects the normal loss of gray matter during adolescence, presumably related to normal processes of synaptic pruning, was exaggerated (Figure 62–6).

The antipsychotic drugs used to treat schizophrenia all act on the dopaminergic pathways of the forebrain. The first effective antipsychotic drug, chlorpromazine, was developed for its antihistaminic and sedating effects and not for its psychiatric effects. Chlorpromazine was later found to be effective in treating the agitation of patients with schizophrenia and manic depressive illness. Based on these acute calming effects, chlorpromazine and many related drugs were initially described as major tranquilizers. By the mid 1960s, however, it became clear that these drugs were not simply acting as tranquilizers but were specifically reducing the positive symptoms of schizophrenia, such as hallucinations and delusions. They were also effective in treating the psychotic symptoms that can occur in mood disorders, such as mania or severe depression.

The antipsychotic drugs had less impact on the negative symptoms of schizophrenia and little or no impact on cognitive deficits. Patients did improve enough to leave the hospital. Indeed, the widespread use of antipsychotic drugs paved the way for the large-scale release of patients with schizophrenia from psychiatric institutions. Unfortunately, these patients did not return to their premorbid level of functioning. The recognition that the sedating properties of early antipsychotic drugs were undesirable side effects led to the development of newer less sedating antipsychotic compounds. In addition, all of the first-generation antipsychotic drugs, with the exception of clozapine, produced Parkinson-like side effects in the extrapyramidal tract such as stiffness, tremor, and difficulty initiating movements.

Because Parkinson disease is caused by the loss of dopaminergic neurons in the midbrain, the occurrence of Parkinson-like symptoms with antipsychotic drug treatment suggested to Arvid Carlsson that these drugs decreased dopaminergic transmission. Following up on this idea, Carlsson established that the antipsychotic drugs block dopamine receptors. Two families of dopamine receptors are known. The D₁ family, which in humans includes D₁ and D₅, are coupled to stimulatory G proteins that activate adenylyl cyclase. The D₂ family, which includes D₂, D₃, and D₄, are coupled to the inhibitory G protein (G_i) that inhibits the cyclase. The D₂ family of receptors has also been shown to signal through an independent pathway involving β arrestin₂ (β-arr₂) and Akt, a protein kinase previously known as protein kinase B. The family of D₁ receptors are expressed in the striatum and are the major type of dopamine receptor in the cerebral cortex and hippocampus, while the D₂ family of receptors is expressed most densely in the striatum, but also in the cerebral cortex, amygdala, and hippocampus. Correlations between receptor binding studies and clinical efficacy in reducing positive psychotic symptoms indicate that the D₂ family is the main target of the therapeutic actions of antipsychotic drugs on positive symptoms (Figure 62–7).

Antipsychotic drugs not only treat acute relapses of schizophrenia and other psychotic disorders, but continuous treatment with these drugs reduces hospitalization because it markedly increases the time between relapses. Unfortunately, the side effects that occur with administration limit their long-term use. A second generation of antipsychotic medications has been developed based on the observation that clozapine has less likelihood of causing Parkinsonian side effects than the other drugs and can also produce therapeutic responses in some patients with schizophrenia for whom other drugs have not worked. (Unfortunately clozapine has other serious side effects that limit its use.)

Based on the properties of clozapine, some of the second-generation drugs were designed to have somewhat lower affinity for D₂ receptors than the first-generation drugs, and some also block the serotonin 5-HT_2A receptors, an action that was thought to protect against motor side effects. Recent large-scale clinical trials of the newer drugs have been disappointing, however, showing little incremental benefit over the older antipsychotic drugs. None of the newer drugs is equal to clozapine in efficacy.

Because drugs that reduce positive symptoms do so by blocking D₂ receptors, investigators have asked: What is the role of dopamine in the symptoms of schizophrenia? Although some drugs that block D₂ receptors reduce psychotic symptoms, other drugs that increase dopamine at synapses (such as amphetamine and cocaine) can produce psychotic symptoms, especially paranoid symptoms. Thus Carlsson suggested that dopaminergic systems are hyperactive in schizophrenia.

The most direct evidence for this idea comes from studies in the mid-1990s that found that amphetamine-produced increases in dopamine release were greater in schizophrenic patients than in healthy subjects. These studies suggest that abnormalities in amphetamine-sensitive processes—such as dopamine storage, vesicular transport, dopamine release, or dopamine reuptake by presynaptic neurons—might lead to hyperactivity in the subcortical dopaminergic systems and could contribute to the positive symptoms of schizophrenia, the symptoms that respond to antipsychotic drugs.

Although dopamine activity might increase in subcortical regions of the brain in schizophrenia, there is also some evidence that it might decrease in cortical regions and that this might contribute to the cognitive symptoms. In particular, the number of D₁ dopamine receptors in the prefrontal cortex is thought to be reduced in schizophrenia, an interesting idea because D₁ receptors have been shown to play a role in working memory and executive functions reliant on prefrontal cortex.

Glutamate, the major excitatory neurotransmitter in the brain, also has been implicated in schizophrenia, albeit indirectly. Phencyclidine and ketamine, which block the NMDA-type glutamate receptor and which were originally developed as anesthetic agents, produce psychotic symptoms. In healthy subjects ketamine also produces cognitive dysfunction that mimics, at least to a degree, the cognitive abnormalities seen in schizophrenia. This has led several investigators to explore the idea that decreased function of NMDA-type glutamate receptors might play a role in producing some of the positive and cognitive symptoms of schizophrenia. These studies indicate that positive and cognitive symptoms are probably the result of abnormalities in several transmitter systems that act either in parallel or in combination with dopamine.

Schizophrenia is a chronic, profoundly disabling disorder characterized by dramatic psychotic symptoms, as well as deficits in emotion, motivation, and cognition. The cognitive deficits impair the ability of people with schizophrenia to regulate their behavior in accordance with reasonable, stable goals. The result is that people with schizophrenia are frequently unable to hold down simple jobs, even at those times when antipsychotic drugs effectively control their hallucinations and delusions.

Once considered a purely psychological reaction to the family environment, it is now clear that schizophrenia is highly influenced by genetic risk factors; indeed, with modern genetic technologies the first convincing risk genes are being identified.

Postmortem studies and neuroimaging are documenting loss of gray matter in the prefrontal and temporal cerebral cortex. Functional neuroimaging is revealing the basis of the disabling cognitive symptoms. Despite this progress, the drugs that we have to treat schizophrenia, as useful as they are, still leave patients seriously symptomatic awaiting new discoveries from neural science.