All psychiatrists think they know what schizophrenia is; yet, the more one sees patients with diagnoses of schizophrenia, the more one questions what statements about the illness can be reasonably defined as a “fact” (
1,
2). The illness has a very heterogeneous prodromal stage and an even more variable outcome. Positive symptoms of auditory hallucinations and multiple, sometime very bizarre, delusions may also be accompanied by agitated and occasional violent behavior as well as severe affective symptoms, such as depression or mania. In rare instances, it is accompanied by catatonic behavior of slow, if any, movements and mutism. Often the diagnosis with all these characteristics becomes labeled as “schizoaffective” or, in some episodes, “bipolar with psychosis.” A patient’s presentation often changes over time, and the lifetime diagnosis may no longer be clear to even the most experienced psychiatrist (
3). Do any of these differences have an underlying brain mechanism?
The underlying basis for all the phenomena characteristic of schizophrenia has been researched for years, and yet no one specific and sensitive diagnostic marker has emerged. Even Emil Kraepelin, when he differentiated dementia praecox from manic-depressive insanity in the early 1900s, put forth pictures of degenerative cells that he thought were present some place in the brain and were responsible for the poor course of illness (
4). As a result, this downward degenerative course has often been described as “Kraepelinian schizophrenia ” in the literature (
5,
6), although it is now known that the Kraepelinian designation does not describe all or even the majority of schizophrenia as it is currently observed (
7). However, over time, the course of illness may also have changed with the initial widespread use in the 1960s of the first-generation antipsychotic medications, the use of which may have had a profound, often unrecognized, effect on the course of illness.
The cause of the underlying brain pathology in the case of schizophrenia is likely genetic or related to gene expression. There are many families that provide examples of such genes at work. For example, Robert Kolker, in his best-selling book
Hidden Valley Road (
8), describes life’s experiences of one such middle-class American family with 12 siblings, six of whom (all male) developed schizophrenia when they were each in their 20s. Although this occurrence may be unusual, it speaks to the problem. That is, how do the genes produce the vast amount of reported neuropathology? How does this neuropathology, in turn, produce the symptoms? Finally, how can the findings from research aid the clinician who cares for patients with schizophrenia? Researchers and clinicians currently do not have enough information to fill those gaps.
In summary, this clinical translational review recognizes the gaps that currently exist between research findings in the brains of people with schizophrenia and the clinical utility of examining the brain among people who may be developing schizophrenia or who have been diagnosed as having the illness. Nevertheless, clinicians who evaluate and treat patients with schizophrenia could benefit from an understanding of how the heterogeneous connections within and among regions of the brain contribute to the symptoms that characterize the illness of each patient and why response to medications may or may not occur. In the future, psychiatric diagnoses may be governed by the changes seen in the brains of patients rather than the clinical symptoms they produce.
Historical Perspective
Joining “bandwagons” that are popular and not being able to “think out of the box” cause stark slowing of progress by years and even generations, and so it has been in schizophrenia research.
If Kraepelin at the turn of the 20th century began this quest for understanding the neuropathology of dementia praecox, Freud unfortunately diffused the matter to the extent that his theories and speculation about child-parent interactions halted progress in the field of neuroscience for many years.
According to the psycho-dynamic approach put forth by Sigmund Freud and his followers, schizophrenia is the result of the disintegration of the ego and its separation from reality (
9,
10). The death of the ego then represented a complete loss of self-identity, which, in turn, led to schizophrenia. This disorder happened, according to Freud, because of an unsuccessful attachment to the parent of the opposite sex (
11), resulting from disordered family patterns.
Freud’s theory of schizophrenia was most explicitly put forth in 1911 in “Psycho-Analytic Notes on an Autobiographical Account of a Case of Paranoia” (the Schreiber case) (
12) and in 1914 in “On Narcissism” (
13). At that time, Freud differentiated between schizophrenia and paranoia and explained his theory according to his concepts of sexual development. In schizophrenia, he postulated a fixation point at the autoerotic stage (just the first 1–2 months); in paranoia, it stopped at the narcissistic phase (immediately following the autoerotic phase). Years later in the lives of individuals with those early fixations, he went on to speculate that the outbreak of schizophrenic or paranoid symptoms would be brought about by a current stress. He formulated that paranoid delusions were based on unconscious homosexual impulses (
14). Although this made for a creative, albeit intricate and complicated, story that the psychoanalytic community was enthusiastic about, the scientific evidence for its basis was lacking.
In 1948, however, it was Frieda Fromm-Reichmann (
15), a follower of Freud and psychoanalyst for schizophrenia, who coined the term “schizophrenogenic mother” when she wrote that “the schizophrenic is painfully distrustful and resentful of other people, due to the severe early warp and rejection he encountered in important people of his infancy and childhood, as a rule, mainly in a schizophrenogenic mother.”
From the late 1940s to the late 1970s, the concept of the “schizophrenogenic mother” was popular in the psychiatric literature; as a result, mothers, and in a broader sense the family, were blamed for the illness that appeared in their offspring. Research later confirmed that the mother who could cause schizophrenia in her offspring did not exist (
16). However, this concept was taught to new generations of psychiatrists by many educators who practiced family therapies and did not understand that hard scientific evidence was lacking.
These theories remained popular, as did psychoanalysis in general, for many years, slowing down progress in biology among psychiatric researchers. Although some pneumoencephalographic studies showed that people with schizophrenia had large cerebral ventricles (
17,
18), this finding was largely unnoticed as evidenced by a lack of citation and replication studies early on. It was not until the much publicized small study in
The Lancet by Johnstone and colleagues (
19) was reported that interest in the biology of schizophrenia took hold. Weinberger, a young enthusiastic thinker at the time, put together all that he observed through computed tomography to conclude that schizophrenia was a neurodevelopmental disorder beginning in utero and manifested in different ways at different times of maturity (
20). After his theory was published in 1987, most researchers agreed it made sense, and very few then were willing to acknowledge that the ventricular enlargement had not yet been proven to be static and could have been progressive over time. This notion of progressive change unfortunately was not consistent with the focus on neurodevelopment and thus was dismissed by many (including Daniel Weinberger, who 30 years later made no mention of the possibility of progression in a review) (
21). Nevertheless, progressive brain change may very well be an important part of the illness pathology, as detailed next.
Current Understanding of the Brain in Schizophrenia
From a clinical perspective, the most important issue is how to study a patient with schizophrenia to provide the best treatment options, outcome, and quality of life. A good clinical history relevant to the brain, such as having had head trauma, would be more useful than initially thought. Gathering a detailed family history helps to understand not only whether a genetic tendency might be prominent but whether the pattern of inheritance gives clues to how penetrant the genes are and whether there might be any of the recognizable genetic syndromes in rare cases, such as the chromosome 22 deletion syndrome (DiGeorge syndrome). A detailed history of childhood developmental milestones, experienced traumatic events, school and social difficulties, as well as seizure or other neurologic conditions can give clues to the onset of pathology. A cognitive battery of tests of memory, executive functioning, and processing speed might give a window into brain functioning without having to order a magnetic resonance imaging (MRI) scan, which can be a difficult procedure for patients to complete.
However, when possible, an MRI at the first episode could be essential, not only to rule out other neurological processes, such as multiple sclerosis, tumors, and so forth, but to provide prognostic indicators and to make faster gains toward a better treatment of choice, such as clozapine or cognitive therapies that may aid in correcting deficits that have emerged in functioning. Nevertheless, it should be noted that the findings reported to be present in MRI scans among patients with schizophrenia from quantitative data in research studies are rarely documented by clinical radiologists who read the scans. No measurements of structures are performed against normative data by clinical radiologists. Unless there are grossly abnormal deviations, the scan will likely be read as normal; therefore, it is only by looking at them individually can a psychiatrist with training in neuroradiology have the sensitivity to know when subtle deviations are present. Having that first scan, however, is the much needed initial assessment that can then be a baseline guide for comparison with a repeat scan in subsequent years for assessing whether progressive changes are taking place.
Finally, it has been well noted that in general, in 21st-century psychiatry and many other disease-related research fields, investigators have dropped the competitive norm to come together on an international level for collaborating and combining data sets. The most important such collaboration in brain imaging is the Enhancing Neuro Imaging Genetics through Meta-Analysis (ENIGMA) network (
https://enigma.ini.usc.edu), directed by Dr. Paul Thompson at the University of Southern California. ENIGMA includes perhaps the most up-to-date and reliable data about the brain in any one of the many disorders and specifically for schizophrenia (
http://enigma.ini.usc.edu/ongoing/enigma-schizophrenia-working-group).
Neurodevelopment in Schizophrenia
It is now known that several genes influence the orchestral events that lead to the synchronized development of the whole human brain, and the manner in which they do this progression is precisely coordinated and timed (
22,
23). How this process is conducted is unknown and one of the amazing complexities of life. There are hints from the literature that very early on subtle deviations are present in these growing brains that subsequently express schizophrenia. For example, pregnancy and birth complications have been long reported to be associated with schizophrenia, including premature births (
24). There were famous home movies taken by parents of their children and collected years later by Walker and Lewine (
25) that showed abnormal movements in babies who had later developed schizophrenia. In addition, developmental milestones (i.e., age when first walking and talking) have been significantly delayed in children who are later diagnosed as having schizophrenia (
26). These indirect observations all suggest an early neurodevelopment origin to schizophrenia. MRI studies of neonates at high risk for schizophrenia, examining cortical thickness, gray matter volumes, and white matter changes, all in regions implicated in studies of adults with schizophrenia, also support a neurodevelopmental abnormality (
27–
29).
Brain Structure in Imaging Studies
Imaging studies of schizophrenia with computed tomography, then later with MRI, have been extensive. Their findings may be summarized as follows: structural changes in both gray and white matter are pervasive in the brains of people with schizophrenia, although these changes are very heterogeneous—that is, structural changes differ among patients, and not all structural changes found are specific to schizophrenia. Furthermore, none of these structural changes are diagnostic of the illness.” (
30–
41) (
Table 1). As researchers and clinicians move toward a more personalized approach to treatment in psychiatry, knowing how any given patient’s brain has changed is helpful so that a rehabilitation program based on specific deficits can be part of the treatment plan for the patient.
Functional Imaging Studies
As a result of underlying brain structural deviations, connectivity between structures will be anomalous, and brain networks will be the ultimate key to symptom formation (
42,
43). Thus, processing of information will be found to be abnormal when studying patients by positron emission tomography (measurement of blood flow or glucose metabolism in specific regions) (
44,
45) or by functional MRI using working memory or language-related stimuli or when just examining the resting state (
46–
50). Several of these studies have been done. In general, they show what is to be expected, that is, people with schizophrenia and also those at high risk for schizophrenia show abnormal brain functioning specifically in frontal and temporal lobe functions, such as working memory, executive functioning, processing speed, and language processing and production (
46,
47,
49,
50). In fact, at rest, when thoughts are possibly randomly going through the brain (described as the default network), abnormalities exist among people with schizophrenia in structures including the medial frontal and posterior cingulate cortices (
51).
Postmortem Brain Studies
From early times (see the Historical Perspective section), postmortem brain collections of people with schizophrenia were considered the focus to understand the biology of the disorder, and several postmortem brain banks have collections of people with schizophrenia (i.e.,
52,
53). However, interpretations of this body of work were hampered by the chronicity of the illness at the time the individual came to autopsy and thus could only be considered an examination of the end stage of illness rather than how it developed. As a result, cause and consequence can never be resolved postmortem. Yet, many interesting findings came out of the notable large collections of brains that were examined in the mid-20th century and beyond. There were also many findings reported that could never be replicated, not surprisingly, given the small sample sizes inherent in postmortem studies. The effects of years of medication on the brain, artifacts of the postmortem procedure, and comorbid medical illnesses were only a few of the other confounds hampering interpretation of these results (
54,
55). However, some of these issues can certainly be overcome if those investigators with brain banks relevant to schizophrenia worked together to establish large international collaborations to examine some of the questions that can be answered only by examining the brain directly (
56,
57).
Neurochemical Abnormalities as Seen Through In Vivo Spectroscopy
Schizophrenia has long been thought to be a consequence of abnormalities in major neurotransmitter systems. The dopamine hypothesis, concerning the underlying neurochemistry of schizophrenia, originated many years ago (
58,
59). The hypothesis has since gone through many refinements on the basis of new scientific evidence. For example, abnormalities in the presence of dopamine are now understood to change over the different stages of illness and are not independent of other neurochemical systems, such as the glutamate, gamma‑aminobutyric acid (GABA), and serotonin neuronal networks (
60).
Magnetic resonance spectroscopy has the advantage of examining an array of neurochemical information and that of specific relevant products of energy metabolism or neuronal activation. It is commonly performed by examining the proton (1-H) nucleus, and depending on the specific MRI sequences, it is also used to quantify
N-acetyl-aspartate, creatinine-phosphocreatine ratios, choline, glutamate, and glutamate-GABA ratios. Although abnormalities associated with schizophrenia from very early on in the illness have been reported, these new studies are ongoing at the time of this writing and have yet to definitively clarify the underlying neurochemical process occurring at all different stages of illness over time (
61,
62). Unfortunately for clinicians, there is no (and there will likely never be) quantitative measure of any neurochemical measure in the blood to be able to support a diagnosis of psychosis because of elevation or deficits in any neurotransmitter. Magnetic resonance spectroscopy holds the best promise, but it needs more refinement in the technology and is only beginning to combine with the other structural and functional MRI studies to understand how neurochemistry contributes to the functional anomalies that, in turn, are expressed in altered brain structural volume.
Progression of Brain Neuropathology
Kraepelin of course first defined schizophrenia as a progressive disorder and described deteriorated brain cells that underlie the disorder (
4). Nevertheless, when Weinberger (
21) publicized how his thinking led him to a neurodevelopmental basis for the illness, the thought that it could be progressive after the onset of symptoms was dismissed. In a painstaking longitudinal study of individuals over time, I and my colleagues (
63–
65) were able to show progressive change taking place in some, but not all, patients. This finding, although confirmed by several subsequent other studies (
66), remains controversial. There have been arguments for and against progressive brain structural change being primary to the process of a schizophrenia illness. Some of the change thought to be primary is hypothesized to be crucial to the timing of brain growth, development, and aging (
67–
69).
Secondary causes attributed to those changes have been thought to relate to MRI technical artifacts (
21) and to the effects of antipsychotic medication (
70). The latter was the most seriously implicated because the drug companies had never provided evidence that this outcome could be a side effect of long-term medication. Yet, some convincing studies seem to show direct effects of these drugs, particularly first-generation antipsychotics, on the brains of nonhuman primates (
71) and then in human studies even with second-generation drugs (
72,
73). This finding stirred a lot of publicity, and clinicians and patients alike interpreted these results to be warnings against the use of antipsychotic medication. However, it needs to be made very clear for clinicians that the benefits of these medications far outweigh the risks. Ultimately, it is clear that the long-term positive effects of antipsychotic medication on the course of the illness are significant, and if withdrawn, the outcomes are negative (
74). Nevertheless, there are observations to doubt iatrogenic reasons for the extensive findings of progressive brain change now in the literature. First, the findings of neuropathology far precede the advent of the first widespread use of antipsychotic medications (
17,
18). Second, before the onset of antipsychotic medications, patients in the prodromal stage are known to have abnormalities that progress in the cases of those who develop schizophrenia (
75–
77). Third, whereas earlier studies emphasized the lack of obvious gliosis as a reason why brain changes could not be progressive in schizophrenia, newer studies using more advanced technology suggest the occurrence of brain inflammatory changes in schizophrenia (
78).Thus, although medications may have some effects on the brain, they do not explain all the findings that have been reported over the years to be present in schizophrenia.
Gaps in Understanding and How They Will Be Bridged
It still is not known how any of the reported neuropathology leads to hallucinations, delusions, and all the primary symptoms of schizophrenia. Speculation includes misconnectivity, so that hearing within the brain is misperceived as coming through the auditory peripheral sensory organs, or things said or happening to people are misperceived to represent other concepts, such as hearing two people saying some words and concluding that they have negative meaning toward the self. None of this conjecture has been replicated or proven in experimental designs, and there remains no proof. In addition, it is unknown what causes all these structural and functional changes, what the timing of the changes is (i.e., when do they begin, and how do they change over time?), whether these changes vary among individuals, and (most important) whether early detection and early treatment prevent these changes and have the ability to alter their course over time. If so, what is the crucial kind of treatment, and is it the same at different time points? How early must clinicians intervene?
All these questions have no answers currently and are the gaps in knowledge about the neuropathology of schizophrenia that have to be filled. Perhaps this uncertainty is why the American Psychiatric Association’s
Practice Guideline for the Treatment of Patients With Schizophrenia fails to mention obtaining an MRI scan at any stage of illness (
79,
80). This guidance no doubt will change over time.
In summary, it has long been understood that schizophrenia is a “brain disease” from both clinical and cognitive windows into the brain and from direct examination of the brain using available instruments of the time. Basic neuroscience has shown the role that genes play in the complex development of the structures and regions of the brain. Currently, even “organoids” of the brain are being grown in vitro in laboratories to be used for investigations of how cells of the brain communicate to constitute what is known as the “human mind.” Moreover, they will enable an understanding of how abnormalities can occur to produce the symptoms that occur in schizophrenia, and, in turn, these results will be targets for new treatments. What a clinician needs to know is that different cortical and subcortical structures have been altered with consequential white matter changes likely in all patients with schizophrenia and that the alterations vary between individuals. Because the extent of the pathology is related to outcome, it is useful to know the status of each patient’s brain. MRI scans will view frontal, temporal, limbic, and other structures that will together give some clues to the functional capacity of each patient and may be helpful for making decisions about the success of the patient in different settings as well as whether to expect recovery of specific functions with antipsychotic medications.