Psychopathology and Psychotherapy

In introducing the scientific study of mind and behavior, we have focused primarily on adaptive behavior, and the normal mental processes that underlie it. There have been occasional references to cases of brain insult, injury, and disease, for the light they shed on normal mental life. Now we wish to examine mental illness in its own right: 

The term psychopathology is, obviously, derived from two Greek roots:

The term makes it clear that mental illness is analogous to physical illness. Just as physical illness involves abnormalities of bodily structure (anatomy) and function (physiology), so mental illness involves abnormalities of mental structure and function -- abnormalities of cognition, emotion, and motivation -- that result in abnormal, deviant behavior.

Defining Psychopathology

It has not proved easy to define psychopathology in the abstract.

By analogy with the concept of pathology in medicine, psychopathology may be defined as abnormalities in mental structures, processes, and states that give rise to abnormal, deviant behavior. But the concept of abnormality implies an opposite construct of normality, from which abnormality deviates. So what do we mean by normality?

Normal mental and behavioral functioning is characterized by:

(Note: I am pretty certain that I derived this list of features of normality from an early edition of a textbook in abnormal psychology written either by Richard Bootzin, or one by Gerald Davison and John Neale, but I can no longer identify the source precisely.)

But having defined a sort of prototype for "normality", what do we mean by deviance?

Deviations from normality can be defined in various ways:

Each of these definitions has certain assets and liabilities. Taken together, these two lists of definitions -- of normality and of deviance -- comprise a kind of "prototype" of the "typical" case of mental illness. Not every mentally ill person will lack all the criteria of normality, or display all the criteria of deviance. But most mentally ill people will display some or most of them, so that the mentally ill are related to each other by a principle of family resemblance.

Syndromes of Mental Illness

In actual practice, mental illnesses are not identified by abstract conceptual definitions of mental abnormality and deviance, but rather in terms of various syndromes characterized by particular signs and symptoms.

The Diagnostic Nosology

I identify nine (9) major categories of mental illness.  Warning: these groupings differ somewhat from such "official" classifications as the Diagnostic and Statistical Manual of the American Psychiatric Association, but the overlaps are clear.

1.  Organic Brain Syndromes,in which there are gross impairments in mental function resulting from known insult, injury, or disease in the central nervous system.

2.  Developmental Disorders,in which there is an abnormal pace of development in one or more mental functions since birth.

3.  Psychoses, in which there are gross impairments in reality testing. Psychoses are often labeled as "functional", meaning that they have no organic cause. However, these disorders are almost certainly "organic" in nature, and as their underlying brain pathology becomes known they may well be shifted to the category of organic brain syndromes.

4.  Neuroses, a set of syndromes that share primary symptoms of anxiety in common. These are also "functional" in nature, but in contrast to the psychoses there is less question of organic involvement; rather, they are commonly attributed to the patient's experiential history of social learning.

5.  Psychosomatic Disorders (also known as psychophysiological disorders) involve actual damage to some organ enervated by the autonomic nervous system, usually associated with psychological stress.

6.  Somatoform Disorders are characterized by physical complaints that have no organic basis. In this respect they are similar to the dissociative disorders, except that the symptoms mimic conditions arising outside the nervous system.

7.  Dissociative Disorders, including conversion disorders, in which there is a disruption of conscious awareness and control.

The dissociative and conversion disorders sometimes mimic the effects of damage to the peripheral or central nervous systems, but in these syndromes there is no evidence of brain insult, injury, or disease.

8.  Personality Disorders (e.g., borderline personality, antisocial personality or psychopathy) are deeply ingrained -- longstanding, inflexible, and pervasive -- patterns of maladaptive behavior which typically develop in adolescence. In contrast to the psychoses and neuroses, whose symptoms are "ego-dystonic" (experienced as alien and unwanted), the symptoms of personality disorders are "ego-syntonic" -- experienced as a part of their normal personality.

9.  Behavioral Disorders consist of specific maladaptive behaviors that occur in the absence of signs of any associated mental disorders (e.g., psychosis, neurosis, or personality disorder).

(10.)  In addition to these forms of mental illness, there are more mundane problems in living (a phrase coined by T.S. Szasz, a famous critic of psychiatry, in his book The Myth of Mental Illness). These include:

These problems don't remotely resemble mental illness, but they can be extremely distressing to the people involved. Accordingly, they are often treated by mental health professionals, including counseling psychologists as well as clinical psychologists, psychiatrists, and clinical social workers.

Culture-Specific Syndromes

Schizophrenia, depression and anxiety disorder, like cancer, heart disease, and measles, are found everywhere -- though their incidence and precise manifestation can vary from culture to culture.  While this mean seem puzzling at first, the existence of culture-specific syndromes only underscores the point that the individual's mind and behavior exist in and are shaped by sociocultural context, which is why psychology is both a biological and a social science.

In addition, there are certain forms of mental illness that are encountered only in particular cultures.  For example:

These syndromes are rarely seen in western developed countries -- except, perhaps, among recent immigrants from these regions. 

An interesting recent case is uppgivenhetsyndrom ("resignation syndrome"), which has been diagnosed among refugee children in Sweden who face deportation -- only in Sweden (at least so far), and only in refugee children (not adults).  These children (and, for that matter, their parents) are under constant, prolonged stress -- first from the conflict that made them refugees in the first place, then from the difficult migration from their home country through Europe to Sweden, and then from the uncertainties of refugee life:  How long will they be able to stay?  How will they live while they are here?  When will they be able to return home?  What will things be like when they get there?  In fact, some refugees are denied asylum, even in a country like Sweden, and it's in these children that uppgivenhetsyndrom is diagnosed.  A typical patient appears to be unconscious, even comatose: “totally passive, immobile, lacks tonus, withdrawn, mute, unable to eat and drink, incontinent and not reacting to physical stimuli or pain.”  However, they are not in a coma.  Their reflexes are normal, as are cardiovascular signs such as pulse rate and blood pressure.  In fact, they show no signs of neurological or any other physical illness.  For this reason, even though these children are obviously under a great deal of stress, uppgivenhetsyndrom is not a stress-related psychophysiological disorder, precisely because there is no evidence of any organic damage -- as you would find in extreme cases of Selye's General Adaptation Syndrome (discussed in the lectures on "The Biological Bases of Mind and Behavior").  In our terms, uppgivenhetsyndrom appears to be a culture-specific form of somatization disorder.  Cases started appearing in the early 2000s, and by 2005 more than 400 cases had been diagnosed.  The children typically recover if their families are permitted to stay in Sweden, especially if they (and their families) also receive psychotherapy aimed at their underlying state of fear and hopelessness (Bodegard, Acta Paediatrica, 2005; see also "The Apathetic" by Rachel Aviv, New Yorker, 04/03/2017, which also discusses culture-specific syndromes in general).  

Cultural differences can also work in reverse, preventing "universal" illnesses from being recognized.  The Hmong people of Laos recognize a condition known as  quag dab peg -- literally, “the spirit catches you and you fall down”; it is treated through religious rituals.  In the West, this same condition is known as epilepsy, and is usually treated quite effectively with drugs.  For an excellent treatment of this problem among Hmong refugees in the Unites States, see The Spirit Catches You and Then You Fall Down: A Hmong Child, Her American Doctors, and the Collision of Two Cultures by Anne Fadiman (1997).

Although there are some culture-specific forms of mental illness, for the most part the major psychiatric syndromes are considered universal.  In Crazy Like Us: The Globalization of the American Psyche (2010), Ethan Watters argues that DSM has become a kind of cultural export, shaping non-Western views of mental illness and its treatment.  Watters suggests that this is a bad thing -- a kind of intellectual colonialism.  On the other hand, nobody complains when other aspects of Western science and medicine are exported to non-Western countries, as in the case of treatments for HIV/AIDS, Ebola, or Zika.  It's likely that schizophrenia, depression, anxiety, and other major mental illnesses are, indeed universal.  But still, non-Western countries likely have something to teach us about prevention and treatment.  For example, epidemiological studies have found that the prospects for recovery from schizophrenia are much better in some cultures than others - -suggesting that, when it comes to mental illness, it's not all in the genes and neurotransmitters.

Structure and Function

In many ways, mental illnesses are analogous to the physical illnesses diagnosed and treated by physicians and other medical professionals. Just as physical illness stems from abnormalities in bodily structure or function -- a weak heart valve, or bacterial infection, or whatever -- so mental illness stems from abnormalities in mental structure or function -- a defect in the system for affect regulation, perhaps, or just acquiring, through learning, some maladaptive belief or expectation.

The Medical Model of Psychopathology

In fact, the language of medicine pervades our discussion of psychopathology. Thus, we have:

Mental illness is diagnosed by

These symptoms and signs of mental illness may be grouped into

These analogies are one aspect of the medical model of psychopathology.

Beyond these analogies, the medical model also has implications for the nature of mental illness. However, these implications are commonly misunderstood. It is commonly believed that the medical model ascribes mental illnesses to organic causes. That every psychiatric syndrome is ultimately an organic brain syndrome. As Ralph Gerard, one proponent of this viewpoint, once put it:

"Behind every twisted thought there lies a twisted molecule".

Similarly, Eric Kandel, the psychiatrist who won the Nobel Prize for his studies of long-term potentiation in Aplysia, discussed in the lectures on "Learning", has stated that "All mental processes are brain processes, and therefore all disorders of mental functioning are biological diseases....  The brain is the organ of the mind.  Where else could [mental illness] be if not in the brain?" (quoted by Kirsten Weir in "the Roots of Mental Illness", Monitor on Psychology, 06/2012). 

This "somatogenic" view of mental illness is quite popular, but it is not what the medical model is about. All the medical model asserts is that mental illness has natural causes.According to the medical model, the causes of mental illness may be biological in nature, or they might be psychosocial in nature. All that matters is that they are natural causes that can be ascertained through the methods of empirical science -- namely psychology and related fields. By extension, the medical model holds that mental illness can be treated and prevented by methods derived from scientific research.

Misunderstanding the Medical Model

However, there are considerable misunderstandings abroad about the nature of the medical model -- including misunderstandings perpetrated by many writers of introductory textbooks in psychology. For example, the 4^th edition Gleitman's Psychology (1995, p. 722), the book that I have used most often in teaching introductory psychology, described the medical model as follows:

Some authors endorse the medical model, a particular version of the pathology model [which assumes that symptoms are produced by an underlying pathology, and that the main goal of treatment is to discover and remove this pathology], that assumes... that the underlying pathology is organic. Its practitioners therefore employ various forms of somatic therapy such as drugs. In addition, it takes for granted that would-be healers should be members of the medical profession.

Many other introductory textbooks (as well as texts in abnormal and clinical psychology) have similar passages. For the most part, they are intended to distinguish an ostensibly somatogenic medical model from the psychogenic models associated with cognitive and behavioral therapy, or to distinguish the profession of psychiatry, with its emphasis on drugs and other physical treatments, from clinical psychology, with its emphasis on behavioral interventions. This common association of the medical model with somatogenic theories and biological treatments reflects a deep misunderstanding, and what I have presented here follows is an attempt to give an alternative perspective on this issue, based on Siegler and Osmond's (1974b) sociological analysis of the medical model,Models of Madness, Models of Medicine (see also Shagass, 1975).

Interestingly, the Osmond of Siegler & Osmond is Humphrey Osmond (1917-2004), a pioneering LSD researcher who (in 1957) coined the word psychedelic to describe the effects of that and other hallucinogenic drugs. Osmond gave LSD to Aldous Huxley, who wrote The Doors of Perception about the experience -- a book from which the rock group The Doors, led by Jim Morrison, took their name. Initially, Osmond thought that LSD would serve as a laboratory model (see below) of schizophrenia (or, at lest, of schizophrenic hallucinations), but he later focused his attention on the potential of it and other psychedelic drugs to treat alcoholism and promote "transcendent" alterations in consciousness (Osmond's obituary appeared in the New York Times, 02/22/04).

According to Siegler and Osmond, the history of psychology can be traced in terms of three major models of psychopathology. The supernatural model prevailed before the 18th century Enlightenment. It assumes that psychology reflects the possession of the individual by demons; by implication, the proper response to psychopathology is exorcism. The moral model, which prevailed in the late 18th and early 19th centuries, assumes that psychopathology -- or, more precisely, abnormal behavior -- is deliberately adopted by the individual, much in the manner of criminal behavior; by implication, the proper response to psychopathology is confinement and other forms of punishment. The medical model, which began to emerge in the 19th century, assumes only that psychopathology is the product of natural causes that can be identified by the techniques of empirical science. By implication, the proper response to psychopathology is diagnosis according to a scientifically validated system, and attempts at cure or rehabilitation by means of scientifically proven methods. Contrary to the popular view, the medical model does not assert that psychopathology is the product of an abnormal biological condition, or that it should be treated only with drugs or surgery. Rather, the medical model is centered on particular rules regulating two primary social roles: the doctor and the patient.

To illustrate the differences between these models, consider the 1973 decision by the American Psychiatric Association, to "de-list" homosexuality as a mental illness. As Charles Silverstein (2011) has noted -- he was one of the psychologists who persuaded the psychiatrists to change their position -- at the time, "homosexuality was considered a crime, a sin, and a mental pathology". So, in that case, homosexuality fell under both the supernatural model (it was considered to be a sin, the work of the Devil), the moral model (it was considered to be a crime, a willful antisocial act), and it was considered to be an illness (a sexual pathology).

The doctor (who does not have to be a physician, or even hold a doctoral degree) possesses a special kind of authority called Aesculapian (after Aesculapius, the Greek god of medicine). Aesculapian authority is a combination of three other kinds of authority recognized by sociologists:sapiential authority, by virtue of the doctor's special knowledge and expertise;moral authority, by virtue of the doctor's concern for the afflicted individual; and charismatic authority, by virtue of the afflicted person's faith that the doctor will be of help. Note that doctors lack structural authority: they cannot enforce their prescriptions, resulting in a markedly low rate of compliance. The doctor's role is to investigate the disorder at hand, by means of procedures that might be unpleasant, intrusive, or even frightening. On the basis of this investigation the doctor makes a diagnosis, informs the afflicted person about the nature of his or her problem, absolves the patient of blame (it is critical to medical ethics that people are not blamed, and thus punished, for their illnesses), and finally creates the conditions for the afflicted person to return to health and his or her proper role in society.

The patient enacts his or her part by taking on the sick role: he or she must seek help from the doctor, and cooperate with treatment; in return, the patient is exempt from some or all of his or her responsibilities during treatment. Note that a doctor's order has supreme authority in society -- it can exempt the person from jury duty, military service, and final examinations. It has this power by virtue of our society's implicit adoption of the medical model and the sick role. However, patients cannot remain in the sick role forever: they must leave it eventually, either by recovering or dying.

A special case is when the illness is chronic, and nothing more can be done to achieve a cure. Under these circumstances the role relationships change. It is the responsibility of the doctor to remove the sick role, and confer the impaired role on the afflicted patient. At this point the patient must leave the hospital and active treatment. What once was an illness is transformed into a handicap; and the doctor is replaced by a rehabilitation specialist. Patients are no longer absolved from their responsibilities: they must return to some socially productive activity, do things for themselves, and cope with their handicaps as well as possible.

What has just been described is what Siegler and Osmond (1974) call the clinical medical model, which is one of many different versions. All versions of the medical model assume that disease is the product of natural causes, and that the proper response is scientifically based treatment. However, they differ in terms of their role relationships. In the clinical medical model, the goal is to cure disease in an individual, and the role relationships are doctor and patient.

So much detail has been devoted to the medical model because it has been subject to so much misunderstanding -- and also because it gives us the opportunity to unite two social sciences, psychology and sociology, at least for a moment. However, the interested reader should reflect on the implications of the medical model(s) for understanding psychopathology -- its nature, causes, treatment, and prevention. And also reflect on the proposition that many of the abuses frequently attributed to mental health professionals -- such as the confinement of mental patients in the back wards of mental hospitals, without any active treatment -- actually represent violations, not expressions, of the medical model.

Excerpted from Kihlstrom, J.F. (2002), "To honor Kraepelin...: From symptoms to pathology in the diagnosis of mental illness". In L.E. Beutler & M.L. Malik (Eds.),Alternatives to the DSM (pp. 279-303). Washington, D.C.: American Psychological Association.

Diagnosis as Categorization

The diagnosis of mental illness is an act of categorization in which patients (or their illnesses) are assigned to categories based on the same feature-matching process we use to categorize other objects.

Formal psychiatric diagnosis essentially began with a French physician, Jean-Etienne Dominique Esquirol (1772-1840), who drew a fundamental distinction between the insane, the mentally deficient (today we use the term intellectually disabled), and the criminal. In the 19th century, Emil Kraepelin (1856-1926) divided the psychoses into two major categories --dementia praecox (early dementia, or what we now call schizophrenia) and manic-depressive illness (what we now call affective disorder). And a little later, Pierre Janet (1859-1947) did for the neuroses what Kraepelin had done for the psychoses, dividing them into two major categories --hysteria (including what we now call the dissociative and conversion disorders) and psychasthenia (including anxiety disorders and some forms of depression).

Since the 19th century, the number of recognized mental illnesses has grown markedly. The first edition of the Diagnostic and Statistical Manual for Mental Disorders (DSM-I), published in 1952, listed only about 100 different syndromes. The latest edition,DSM-V, published in 1994, listed almost 300.DSM-VI is due to be published sometime around 2012, and we'll see how many mental illnesses there are then!

Whatever its edition,DSM is essentially a catalog of mental illnesses, with a list of the symptoms characteristic of each. Diagnosis is essentially a feature-matching process that asks whether a patient has the symptoms associated with a particular syndrome or disease. In other words, diagnosis is an act of categorization, so it is interesting to look at what kind of categories the diagnostic categories are.

Diagnostic Categories as Proper Sets

In the past, the diagnostic categories of mental illness were at least tacitly construed as proper sets, where sets of symptoms served as defining features of a syndrome, singly necessary and jointly sufficient to define an illness (such as schizophrenia) or a person (such as a schizophrenic) as having some illness.

In this way, the traditional psychiatric nosology formed a conceptual hierarchy with superordinate categories (organic vs. functional, psychotic vs. neurotic) at the top. Subordinate categories were then created by adding symptoms (such as loss of reality testing or problems with anxiety) as defining features.

The Case of Schizophrenia

The nature of traditional psychiatric diagnosis, as a perfectly nested hierarchy of proper sets, is exemplified by the work of Eugen Bleuler, a Swiss psychiatrist who in 1911 redefined dementia praecox as "the group of schizophrenias".

Bleuler clearly construed the 4 As as defining features of schizophrenia:

The boundaries between schizophrenia and manic-depressive illnesses were clear: you could have one illness or the other, but not both. And the boundaries between schizophrenic subtypes were also clear: you could be hebephrenic or catatonic, but not both.

A similar hierarchy of syndromes developed around manic-depressive illness.

Problems with the Diagnostic Categories

This view of the diagnostic categories as proper sets, recognized by symptoms that were singly necessary and jointly sufficient to define the diagnosis. And early editions of the DSM were at least implicitly structured around this conceptualization, in terms of the "textbook cases" it used to characterize each syndrome.

However, the traditional view quickly encountered problems of a sort that are familiar from the critique of the classical view of categories as proper sets (discussed in the lectures on Thought and Language).The simple fact was that very few patients actually resembled the textbook descriptions of the various syndromes.

Psychiatric Syndromes as Fuzzy Sets

Accordingly, for the third edition of DSM (DSM-III), published in 1980, the diagnostic system was reformed to take into account a new understanding of the structure of natural categories offered by cognitive psychology. Under this "revisionist" view:

Consider, for example, several prominent psychiatric diagnoses, as listed in DSM-5 (2013):

A statistical technique called network analysis can be used to show the relations among the various categories of psychopathology. Denny Borsboom, a Dutch psychologist, and his colleagues created a simple spreadsheet showing the co-occurrence of all the symptoms listed in DSM-IV -- that is, how many times insomnia and fatigue are listed as characteristic symptoms of the same syndrome (such as sleep disorder or depression). The obtained a graph in which each symptom is represented by a node, and nodes are connected whenever two symptoms are characteristic of the same disorder. They then colored each of the nodes, according to the major category of mental illness in which each symptom occurred. The resulting graph shows the extent of symptom overlap in DSM-IV. in which

Regardless of diagnostic category, there was no expectation that all of the symptoms listed under a syndrome will be present in any particular case. In principle, we can observe all possible combinations of characteristic symptoms. Diagnosis, then, is a matter of judgment under uncertainty. Studies of the diagnostic process by Nancy Cantor and her associates show that the certainty with which a diagnosis of schizophrenia is made, for example, will be a function of the number of the patient's symptoms that are highly characteristic of schizophrenia, and the number of symptoms that are more characteristic of other diagnostic categories. In this system, "textbook cases" serve as category prototypes, and there is explicit recognition of heterogeneity among actual patients.

DSM-5

The Diagnostic and Statistical Manual for Mental Disorders  (DSM) was first published in 1952, with a second edition appearing in 1968.  Both manuals were as much literary productions as scientific ones.  They were heavily influenced by Freudian psychoanalysis, centered on the distinction between psychoses and neuroses, and didn't really include clear criteria for making various diagnoses.  All that changed with DSM-III in 1980 (a revised edition, known as DSM-II-TR, for "Text Revision", came out in 1987), and DSM-IV in 1994. In these editions, serious effort went into producing checklists of symptoms by which the various disorders could be reliably diagnosed.

The fifth edition of DSM, known as DSM-5, retains the "fuzzy set" structure of the diagnostic categories, while sometimes revising the criteria for specific disorders.

However, DSM-5 abandons certain other features of previous editions.

Aside from Axes I and II, none of these dimensions got much use in real-world, everyday diagnosis, and so they were dropped.

Some of the changes in DSM-5 are highly controversial.

Some commentators have expressed the fear of diagnostic inflation -- that these kind of changes may make it possible to diagnose anyone with a mental illness -- or, at least, that DSM threatens to "pathologize normality" by classifying normal mental states, like bereavement or even shyness (which can resemble depression), as mental illnesses.  (Benjamin Wolman, a leading psychoanalytic psychotherapist, once wrote a book entitled Call No Man Normal.)  Over-diagnosis, of course, inevitably leads to over-treatment.

In addition, some of those responsible for formulating the DSM-5 criteria had unacknowledged links with the pharmaceutical industry, raising the possibility that increasing the number of people who can be diagnosed with some form of mental illness, however mild, will effectively expand the market for psychiatric medications.  At the very least, these affiliations might have biased the Manual developers toward biological diagnoses and treatments.  At the worst, it raises the possibility that Big Pharma might encourage clinicians to formulate new diagnoses that expand the market for available psychotropic drugs -- what Marcia Angell, the former editor of the New England Journal of Medicine, has called a "patent-extending game" (see her book, the Truth About Drug Companies: How They Deceive Us and What We Can Do About It, 2004). 

Other commentators worry that many DSM-5 diagnoses simply lack validity to begin with.  In an important early paper, Robins and Guze (1970) outlined the criteria for establishing the validity of any diagnostic category:

By these standards, more than 40 years later, many DSM diagnoses do not fare too well.  For example, the American Psychiatric Association proposed to "field-test" DSM-5 before its actual publication, to make sure that clinicians could use it reliably to make diagnoses.  Unfortunately, DSM-5 failed its field trials, in that the rate of disagreement between two clinicians, applying the same criteria to the same patient, was unacceptably high.

This state of affairs has led some investigators to propose alternatives to the DSM.

For now, though, DSM-5 is what we've got.  The Social Security administration won't pay disability benefits, insurance companies won't pay for treatment, and courts won't consider the insanity defense, in the absence of an official psychiatric diagnosis, and DSM is how psychiatrists officially diagnose patients.  It's as simple as that.

Research Domain Criteria

One possible alternative diagnostic system, embraced by the National Institute of Mental Health, is the Research Domain Criteria (RDoC).  Instead of employing traditional diagnostic categories, such as schizophrenia and bipolar disorder, RDoC classifies mental processes into several domains, constructs, and sub-constructs, with each construct and sub-construct hypothetically linked to a different neural circuit in the brain. 

Here is a list of the domains and sub-domains, as listed in a document published by NIMH in 2011.

Following the medical model, it's important to get beyond surface symptoms and signs to underlying pathology.  DSM-5 doesn't do that, it's diagnoses are solely based on signs and symptoms.  Diagnosis in the rest of medicine is of diseases, and the RDoC is a step in this direction.  At the same time, we can worry about the underlying assumption, which is that each element of underlying pathology is linked to a dysfunction in some brain circuit (Insel, "Faulty Circuits", Scientific American, April 2010).  It's not at all clear that a faulty brain circuit must be involved in every instance of mental illness.  Moreover, the identification of such circuits is years, decades, maybe centuries away.  The whole project of identifying the circuits of the brain, which is the raison d'etre of the Human Connectome Project announced by President Obama in 2013, begins with extremely simple nervous systems, such as aplysia (the sea slug studied by Eric Kandel and others), and then move up to the mouse, and only later to the human -- at which point the process of linking neural circuits to the various domains can begin.  Not in my lifetime, nor yours. 

In the meantime, diagnosis needs to move beyond signs and symptoms to the identification of underlying pathology through laboratory tests -- underlying psychopathology.

For now, though, and for the foreseeable future, psychiatric diagnosis is going to be based on symptoms and signs, and organized by something very much like DSM-5.  This line of research in experimental psychopathology has a long and distinguished history, going back to Emil Kraepelin's studies of schizophrenia using Donders's reaction-time methodology.  While the RDoC claims to be "agnostic" about the traditional diagnostic categories (actually, far from being agnostic, it rejects them), experimental psychopathology takes them as a starting point, and tries to identify the pathological mental processes that underlie them.  One result of this work should be the more fine-grained classification, moving beyond superficial symptoms and signs, that the RDoC seeks -- but without the excessive biologizing.

For critiques of psychiatric diagnosis and DSM, see: The Selling of DSM (1992) and Making Us Crazy (1997) by Herb Kutchins. They Say You're Crazy (1996) by Paula Caplan. "The Epidemic of Mental Illness: Why?" and "The Illusions of Psychiatry" an essay-review of several books critical of psychiatry by Marcia Angell (New York Review of Books, June 23 and July 14, 2011). "Head Case: Can Psychiatry Be a Science?" by Louis Menand (New Yorker, March 1, 2010). The Book of Woe: The DSM and the Unmaking of Psychiatry (2013) by Gary Greenberg, who doesn't seem to like diagnosis at all (he also wrote another book, entitled Manufacturing Depression). Saving Normal: An Insider's Revolt Against Out-of-Control Psychiatric Diagnosis, DSM-5, Big Pharma, and the Medicalization of Ordinary Life (2013) by Allen Frances (who led the task force that prepared DSM-IV, but thinks that DSM-5  takes an approach to diagnosis that has become outmoded). "Three Approaches to Understanding and Classifying Mental Disorder: ICD-11, DSM-5, and the National Institute of Mental Health's Research Domain Criteria (RDoC)" by Lee Ana Clark et al. (Psychological Science in the Public Interest, 2017).  A comprehensive overview of contemporary approaches to psychiatric diagnosis, emphasizing differences between DSM and RDoC.

Experimental Psychopathology

Current diagnostic practices classify illness based on surface symptoms, but the symptoms are not the disease. Rather, symptoms are assumed to be caused by some underlying disease process --pathology.

Beyond Surface Symptoms to Underlying Pathology

In medicine, symptoms are not the disease, and progress in medical understanding is marked by a shift in focus from symptoms to underlying disease processes -- to underlying pathology revealed by laboratory analyses of structure and function whose results are interpreted in the light of a theoretical understanding of normal function.

Consider the example of fever, manifested by chills (a symptom about which patients complain) and increased body temperature (a sign, indicated by a thermometer). Well into the 19th century the medical nosology included a large number of different "types" of fever, based on the symptoms accompanying the fever and the circumstances under which the fever occurred. In fact, Benjamin Rush, the pioneering American physician of the Revolutionary War era, thought there was only one disease -- fever.  But nobody diagnoses and treats fever anymore, because fever is viewed as a symptom of an underlying bacterial infection. Physicians may try to reduce a patient's fever, as an emergency measure, but most of their efforts are devoted to identifying the underlying infection (through blood tests) and then treating it (through antibiotics).

In medicine, pathology affects bodily functions through lesions in anatomical structures (e.g., a skull fracture or lung cancer), abnormalities in physiological functions (e.g., hypertension or sinus arrhythmia), or infection by micro-organisms (e.g., the influenza virus; in addition to viral infections, there are also infections by bacteria, fungi, chlamydiae, rickettsiae, and microplasmas). Modern medical diagnosis is based on evidence of these pathological conditions, as revealed by objective laboratory tests.

There are parallels in the medical model of psychopathology:

In medicine, underlying pathology is revealed by laboratory analyses interpreted in the light of our understanding of normal structure and function. In psychopathology, such laboratory research is known as experimental psychopathology. Experimental psychopathology tries to identify the causes of the patient's manifest symptoms. It is basic research on psychopathology, as opposed to applied research on diagnosis and treatment, and it comes in two major forms:

Attentional Deficits in Schizophrenia

Studies of psychological deficit in schizophrenia have a long history, dating back to experimental work by Kraepelin, in Wundt's laboratory, who used Donders's reaction time technique to measure the speed of mental processes in patients with dementia praecox (presenile dementia), the syndrome later renamed schizophrenia. One highly prominent theory of schizophrenia holds that the psychological deficit underlying the patient's presenting symptoms affects the attentional system. According to this theory, schizophrenics suffer from a breakdown in selective attention that renders them highly distractable and unable to filter out irrelevant ideas. The patient shows thought and language disorder because he cannot keep track of what he is thinking and saying; he withdraws socially to shut out this chaos of stimulation. And, in fact, laboratory tests confirm that patients with schizophrenia have a number of difficulties with attention. 

The information-processing deficits in schizophrenia appear to begin at the very beginning of the information-processing sequence.  Think of the multi-store model discussed in the lectures on memory.  In that model, stimulus information is briefly held in modality-specific sensory registers, from which it is extracted into short-term or working memory for further processing, and then encoded into long-term memory.  Studies of information-processing in schizophrenia have focused on the sensory registers, working memory, and attention.  If things go wrong at these earliest stages of information processing, lots of other things will go wrong as well. 

Consider, for example, a study by Wishner and Wahl (1974) of dichotic listening in schizophrenia (there are other studies of this topic, but I chose this one because Wishner was a teacher of mine in graduate school and Wahl was a graduate-student colleague).In this paradigm, different messages are played through earphones to the two ears, and the subject is instructed to shadow one message, repeating its words aloud as they are spoken, and ignore the other. There is little or no impairment when normal subjects shadow a single message, with no distracting message coming over the other ear. When the distracting channel is added, however, subjects begin to make shadowing errors: omissions of parts of the target message and intrusions of the irrelevant message, and recall of the target message.Wishner and Wahl found that schizophrenics were particularly prone to make shadowing errors: more omissions and more intrusions, especially when the target message was played at a relatively fast speed;, and poorer recall of words from the target message, even when no distractor was present. These results are consistent with the hypothesis that schizophrenics suffer from an attentional deficit.

Comparable findings were obtained in a study of visual attention by Saccuzzo and Shubert (1981) employing the backward masking paradigm. In this procedure, an array of digits or letters is presented very briefly, so that there is not much opportunity for it to register in a very short-term memory store known as iconic memory (see the lecture supplements on Memory). Presentation of this array is followed by a "masking" stimulus that effectively displaces the array from iconic memory, preventing it from being further processed in "primary", "short-term, or "working" memory. Therefore, identification of the elements in the array requires highly focused attention, so that the subject can process the information from the array into primary memory before its representation disappears from iconic memory. In the experiment, subjects had to search for the letter T in an array of As, or for the letter A in an array of Ts: if it is present, they simply say "yes" and another trial begins.The investigators varied the "stimulus-onset asynchrony" (SOA), or the interval between the onset of the array and the onset of the mask -- essentially, the amount of time that a representation of the array resides in iconic memory. They found that schizophrenics were generally poorer at target detection than controls, especially at longer SOAs. They concluded that it takes longer for schizophrenics to transfer information from iconic to primary memory. By virtue of this slower rate of information processing, schizophrenics miss a lot of what goes on around them, and are more distractable.

In recent years, much attention has focused on working memory in schizophrenia.  You’ll remember from an earlier lecture that working memory enables an individual to maintain information in an active state, over short periods of time, while it is being manipulated by other cognitive processes.  This information may be extracted from perception or retrieved from memory.  As such, working memory is critical for a wide variety of cognitive processes, such as selective attention, including both focusing on needed information and inhibition of irrelevant information.  A problem in working memory can permeate far into cognitive function, affecting memory, reasoning, problem-solving, and language.    10 

Working memory is often studied with variants of the Sternberg task, discussed at length in the lecture on methods and statistics.  In the Sternberg task, a subject is asked to memorize a small set of items, and then to search that set for a particular target. It’s different from the Sperling task.  In the Sperling task, subjects have to inspect a study set presented as a visual array.  In the Sternberg task, they have to hold a representation of the study set in working memory while they search it.  Sternberg’s basic finding, you’ll remember, was that response latency varied with the size of the search set, indicating a serial, self-terminating search process.  An experiment by Paul Metzak and his colleagues compared schizophrenic patients with a group of normal control subjects who had been matched with the patients on such demographic variables as age and education.

Like Sternberg, Metzak found that accuracy decreased, and response latency increased, with the size of the study set.  But this variable had a greater effect on the performance of the schizophrenic patients than on the normal subjects.  Performance especially deteriorated at the largest set size.  It’s as if the patients were overwhelmed by the information they had to process – maybe because they’re processing information so slowly, as in the Saccuzzo experiments in iconic memory.

Working memory consists of several different components.  First, there are modality-specific buffers which maintain information in an active state.  These elements are similar to the traditional concept of short-term memory, and there appears to be no problem with them in schizophrenia.  Then there is a central executive, which actually guides various information-processing tasks, manipulating and transforming information held by the buffers.  Here is where the psychological deficit in schizophrenia appears to be located.  Other research shows that there is a particular problem with a component of the central executive that represents and maintains contextual information that is relevant to current tasks – information about the task being performed, other information that has been recently processed, and what’s coming next.  This central executive appears to be mediated by the dorsolateral prefrontal cortex.  Interestingly, this region of the brain is part of a system modulated by dopamine, and the antipsychotic drugs used in the treatment of schizophrenia are antagonists of this neurotransmitter.    

Studies of dichotic listening, backward masking, and working memory illuminate schizophrenic deficiencies in central mechanisms of attention, but attention is also regulated by peripheral mechanisms -- as when we turn our heads or eyes to shift our attention from one object to another. Holzman and his colleagues have studied the peripheral mechanisms of attention with an eye-tracking paradigm in which subjects are asked to hold their head in a fixed position and move their eyes to track a target moving smoothly across a screen; they then examine the subjects' pursuit eye movements (PEMs) as they follow the target. Some aspects of these eye movements are not consciously perceptible to the subject, but can be recorded by an electrooculogram (EOG) similar to that used in sleep research, or by special infrared devices.

Holzman et al. have found that in contrast to the smooth pursuit eye movements (SPEMs) characteristic of normals, schizophrenics show eye movements that are more jagged.About 70% of schizophrenic patients show anomalous PEMs, but PEM anomalies also show up in relatives of patients who are not themselves diagnosed with schizophrenia.Accordingly, Holzman has suggested that abnormal PEMs might be a biological marker for schizophrenia, perhaps indicating an underlying malfunction in the frontal-lobe of the brain, particularly in those centers involved in the peripheral control of attention. Interestingly, Holzman and his colleagues have found that abnormal PEMs are related to various aspects of thought disorder, as measured in patients' verbal protocols. However, abnormal PEMs are associated with thought disorder across a wide variety of diagnoses, so this aspect of attentional deficit may not be specific to schizophrenia.

Most experimental studies of schizophrenia focus on various aspects of cognitive function, such as attentional deficits, but schizophrenia can involve problems with emotional function as well. Consider anhedonia, one of the classic "Four As" of schizophrenia. Many schizophrenics show flat or bunted affect, or else affect that is inappropriate to the situation. But that's their display of affect. Recall Lang's multiple systems view of attention, according to which emotional responses have three different components: a subjective feeling state, overt behavior, and covert physiological response.

In one experiment, Kring and Neale (1998) showed emotional (positive and negative) and neutral films to schizophrenic patients and controls, and measured all three components of emotional response. In terms of facial expressions, schizophrenics were, indeed, less reactive than controls -- in terms of the frequency, intensity, and duration. But they were actually more reactive when their covert physiological responses were measured by a psychophysiological index known as skin conductance. The self-reported emotional responses of schizophrenic patients were somewhat muted, compared to controls, but both patients and controls showed the same pattern of response to positive and negative films. So, schizophrenics do not appear to be emotionally responsive, in terms of their facial expressions. But in terms of their subjective feeling states, and their physiological response to emotional stimuli, they appear to be experiencing emotion -- even if they aren't showing it in their behavior.

Attention-Deficit Disorder

The problem with diagnosing mental illness based on symptoms is illustrated by attention deficit hyperactivity disorder (ADHD), which is typically diagnosed in children based on such symptoms as failure to pay attention in school or play, running around the room, climbing on things, fidgeting and squirming, and the like. But adults with ADHD don't necessarily do these things. Instead, they display other, more age-appropriate symptoms such as difficulties in "wrapping up" the final details of a project, getting and keeping things in order, remembering appointments, and the like. Both sets of symptoms may have a common origin in some attentional dysfunction, and perhaps a common standard for diagnosis, applicable to children and adults alike, could be achieved by the development of laboratory tests that assess attentional functions directly.

Laboratory Models of Psychopathology

In addition to studies of psychological deficit, it is sometimes possible to create laboratory models of psychopathology, inducing in normal individuals a syndrome that, in at least some respects, resembles some form of actual mental illness. Laboratory models are rarely if ever exact replicas of mental illnesses, down to the last detail; rather, they usual mimic one or more characteristic symptoms of some syndrome. In a sense, laboratory models constitute theories of how symptoms arise in actual patients, because they are based on the assumption that the causative agents that effectively produce symptoms in the lab parallel those that are present in the real world outside the laboratory. Thus laboratory models can be used to test proposals concerning the origins, treatment, and prevention of mental illness. As such, they can be evaluated on a number of dimensions:

Anxiety Disorders

Like studies of psychological deficit, laboratory models of psychopathology have a history that goes fairly far back in time -- but this time, to Pavlov's laboratory rather than Wundt's. In some early studies of discrimination learning, dogs were conditioned to salivate to a circle or an ellipse, and then the axes of the stimulus were progressively changed, so that the circle became more elliptical, or the ellipse more circular. The result was that, at some point, the dogs became distressed -- seemingly anxious -- a phenomenon that became known as experimental neurosis. One explanation (proposed by Sue Mineka and myself) was that this increase in anxiety occurred because the animals could no longer predict the onset of the food US. Unpredictability causes anxiety (hold this thought, because in a little while we'll discuss another laboratory model that indicates that uncontrollability causes depression).

In addition, conditioned fear has served as a laboratory model of phobia, while conditioned avoidance has served as a laboratory model for obsessive-compulsive disorder.

For a personal account of OCD, as well as good coverage of the scientific literature, see the Man Who Couldn't Stop: OCD and the True Story of a Life Lost in Thought by David Adam (2015).  Adam himself fell into the clutches of OCD after having unprotected not-quite sexual intercourse as a college student.  He became obsessed with the idea that he might contract HIV-AIDS, and went to great lengths to ward off that outcome.

Learned Helplessness and Depression

Of particular interest is the development of learned helplessness as a laboratory model of depression. As discussed earlier (in the lectures on Learning), learned helplessness was initially observed in studies designed to test Mowrer's "two-factor" theory of avoidance learning. In avoidance learning, a conditioned stimulus (such as a tone) is followed by prolonged shock as an unconditioned stimulus. If the animal makes a certain conditioned response (such as moving from one side of the shuttlebox to the other) after the onset of the US, it can turn the US off; termination of the US constitutes reinforcement of an escape response. If it makes the CR after the onset of the CS, but before the onset of the US, the US is prevented from occurring at all; this reinforces an avoidance response. Martin Seligman and his colleagues discovered that prior exposure to inescapable shock interfered with escape and avoidance learning. In their analysis, prior experience with inescapable shock taught the animal that shock was uncontrollable, and this learning generalized to the new situation of the shuttlebox. The learned helplessness experiments underscore the principle that in instrumental conditioning the organism is learning to control its environment.

The animals in these experiments learned to be helpless, but Seligman and his colleagues also observed that they looked and behaved as if they were "depressed" (if you've ever seen a sad dog, you know what they meant). This led Seligman to propose that learned helplessness is a laboratory model for some forms of depression -- i.e., that some people become depressed by virtue of a history of uncontrollable aversive events in their lives.

Subsequent research by Seligman and others, including Steven Maier, Lyn Abramson, and Lauren Alloy, identified a number of parallels between learned helplessness and depression:

The learned helplessness model of depression, as originally formulated, is not complete, and nobody claims that LH underlies all forms of depression. But Abramson and Alloy have suggested that a certain type of depression, which they label helplessness depression, is caused by the kinds of experiences modeled in the LH experiments.

Abramson, Alloy, and their colleagues subsequently modified Seligman's theory with the hopelessness theory of depression. They argued that the experience of uncontrollable aversive events was not enough to make people (or dogs, for that matter) depressed. Sometimes, people (and dogs) respond to uncontrollable aversive events with anger, instead.

Abramson and Alloy proposed that uncontrollability led to depression only when the individual made a certain causal attribution concerning the uncontrollability. They argued that the explanations that people make for various events vary on certain dimensions:

They proposed that uncontrollability causes depression only when the individual makes an internal, stable, global attribution for the helplessness -- as if to say, I can't control this thing, it's my fault that I can't control it, I can never control this or anything else. If you thought like this, you'd get depressed, too, and that's the point.

Abramson and Alloy pointed out that, in contrast to non-depressed people, depressed people are often starkly realistic about their inability to control events -- a characteristic of depressive realism that they contrasted to the illusion of control that is characteristic of non-depressed thought. That is, non-depressed individuals often have an unrealistically elevated sense of control (which is why many of us think we can control chance events by picking "lucky" lottery numbers), while depressed individuals are often quite realistic about the prospects.

Abramson and Alloy identified this pessimistic attributional style as a risk factor for what they called hopelessness depression. They and their colleagues also constructed a personality questionnaire, the Attributional Style Questionnaire (ASQ) to identify people who might be "at risk" for depression, based on this aspect of cognitive style.

Seligman, for his part, took off from his studies of helplessness and depression to focus on the other side of things, and proposed that positive psychology should focus on the sunny side of life, and the positive characteristics that enabled people to be resilient in the face of unpleasant circumstances. In particular, Seligman has proposed that giving experiences of control to children who are at risk for depression (by virtue of their pessimistic attributional style) may help these children avoid actual episodes of depression.

In the domain of the psychoses,amphetamine psychosis can serve as a laboratory model of schizophrenia. High doses of amphetamines, which increase dopamine activity in the brain, lead to psychological symptoms similar to those found in acute schizophrenia -- hallucinations, thought disorder, and paranoid delusions. The behavioral parallels between amphetamine psychosis and schizophrenia is one source of the dopamine hypothesis of schizophrenia, discussed below.

Hypnosis and "Hysteria"

While most laboratory modeling has focused on various anxiety disorders, some investigators -- again, beginning in the late 19th and early 20th century -- noticed a phenotypic similarity between some of the phenomena of hypnosis (e.g., suggested blindness, deafness, and analgesia; suggested paralysis; posthypnotic amnesia) and some of the characteristic symptoms of "hysteria", a cluster of syndromes now known as the dissociative and conversion disorders.

This has suggested that understanding the mechanisms of hypnosis might help us to understand these forms of mental illness as well. In fact, it's been proposed that both the dissociative and conversion disorders result from a division of consciousness that prevents certain percepts and memories from being represented in conscious awareness.

Linking Laboratory Models to Psychological Deficits

Sometimes, laboratory models and studies of psychological deficit go hand-in-hand. Consider experimental research on psychopathy, or antisocial personality disorder. One feature of psychopathy is that these individuals tend not to be responsive to aversive stimulation. On experimental tests, psychopaths often show a failure of avoidance learning, and they also show a failure to respond to punishment. Gorenstein and Newman (1980) observed a similar pattern of behavior in laboratory rats who had surgical lesions in a subcortical area of the brain known as the septum. Rats with septal lesions do not freeze when they are punished, they have difficulty with passive avoidance learning (i.e., learning not to do something in order to avoid punishment), and they also have difficulty with delay of gratification. All of these phenomena, of course, closely resemble the characteristic symptoms of psychopathy. This has led to a theory that, by virtue of some kind of brain disorder, psychopaths, like septal rats, are unable to suppress habitual behaviors in order in order to avoid the aversive consequences of these behaviors.

Experimental research by Newman and others have further clarified the psychological deficit in psychopathy to problems linking attention to the reward system.  In one experiment, subjects played a computerized card game: if they turned over a card they would gain a point if it was a face card, and lose a point if not.  The deck was arranged so that 9 of the first 10 cards were face cards, then 8 of the next ten, 7 of the next 10, and so on, and they could stop whenever they wanted.  Of course, with the deck arranged in this manner, the subjects began accruing lots of points, and then started losing.  Normal subjects generally stopped playing after about 50 cards, while they were still ahead; but the psychopaths continued playing, losing everything they had won -- and more.

Newman's latest theory of psychopathy implicates attention rather than reward per se.  The general idea is that psychopaths have difficulty updating working memory with new information, once their attention has been engaged.  They can focus attention, but they can't disengage and shift it very easily.

Still, there does seem to be something missing in psychopaths' emotional lives.  They don't accurately pick up on other people's facial and vocal expressions of emotion, especially fear; don't respond normally to words with positive or negative emotional connotations.

Returning to Newman's septal rats, Kent Kiehl (2006) has used fMRI to record brain activity in psychopaths during various types of tasks.  He has found deficits in a a system of subcortical brain areas known as the paralimbic system:

For more on psychopathy, see "Inside the Mind of a Psychopath" by Kent A. Kiehl and Joshua W. Buckholtz, Scientific American Mind, September-October 2010.

Linking the Laboratory and the Clinic

We're going to get back to the clinic in a moment, but the point of this whole exercise is that the clinical enterprise of understanding and treating mental illness is not divorced from the sort of basic research that goes on in university laboratories. We can use various research paradigms of the sort discussed earlier in this course -- classical and instrumental conditioning, dichotic listening, and the like -- to understand the basic mechanisms by which mental illness occurs -- as well as ways in which we might more effectively treat mental illness and prevent it from occurring in the first place.

By virtue of basic laboratory research we can move beyond the surface signs and symptoms of mental illness to understand its underlying pathology. And by better understanding its underlying pathology, we will be able to formulate better theories of the causes and cure of mental illness, and better tools for diagnosis, treatment, and prevention.

In fact, laboratory research helps us to identify two different ways that mental illness can occur.

The Biology of Mental Illness

Where mental illness appears to reflect maladaptive social learning, we generally assume that the architecture of the individual's basic mental structures and processes is largely intact, as are the neural substrates of that mental architecture. However, where mental illness appears to reflect an underlying psychological deficit, there are good reasons to think that the neural substrates of that mental architecture are malfunctioning as well -- that is, that the mental disorders are ultimately neurological disorders. This idea is expressed in Ralph Gerard's old adage that

The idea that mental illness have underlying biological causes goes back at least as far as the 19th century -- which is to say, it is almost as old as scientific medicine and scientific psychology themselves. In fact, the history of psychiatry and clinical psychology may be characterized as a cycle in which prevailing views alternate between "somatogenic" theories that mental illness is due to biological causes (i.e., brain insult, injury, or disease) and "psychogenic" theories that mental illness is due to environmental causes, and that the biology of the nervous system is no more relevant to mental illness than it is to normal mental and behavioral functioning.

From Somatogenesis to Psychogenesis and Back Again

The earliest scientific theories of mental illness were neurological theories, based on the assumption, mostly unproven, that patients' symptoms were due to lesions or infections affecting brain tissue. 

With the emergence of Freudian psychoanalysis, in the late 19th and early 20th centuries, the predominant theory of mental illness shifted from somatogenic to psychogenic.  Put briefly, Freud and his followers taught that mental illnesses, particularly the neuroses, had their origins in conflict and defense.  Psychoanalysis, both Freudian and neo-Freudian, dominated psychiatry well into the 1950s.

Another important influence on American psychiatry was Adolph Meyer, who argued that mental patients' problems had their origins in their life histories, not their biology.

The pendulum began to shift back toward somatogenesis in the 1950s, with the introduction of the first psychotropic ("mind-moving") drugs: Thorazine (1954), a "major tranquilizer" used in the treatment of schizophrenia; Miltown (1955), a "minor tranquilizer" used to treat anxiety; and Marsilid (1957), a "psychic energizer" used in the treatment of depression.  These drugs were discovered more or less accidentally, but it was soon learned that they (and other drugs like them) altered the levels of certain neurotransmitters in the brain.  This led to the hypothesis that schizophrenia, depression, and other forms of major mental illness were caused by abnormal levels of these substances. 

Let us just note, in passing, that the logic of this inference is far from airtight.  In fact, it's a variant on the logical error of denying the antecedent, discussed in the Lecture Supplement on Thinking.  The logic seems to go something like this.

The problem is that the efficacy of a treatment says nothing about the cause of the illness.  Nobody thinks that a lack of aspirin causes fever to occur.

And, as it happens, the evidence for the dopamine theory of schizophrenia, and similar theories, is not completely convincing.  There has never been any convincing demonstration that, prior to treatment, schizophrenics or depressives actually suffer from any kind of chemical imbalance in their brains.  Still, theories about chemical imbalances remain very popular.  

In fact, over the past 100 years or so, biological causes have been uncovered for a number of mental illnesses, whose biological causes were unknown at the time they were described, and which had previously been attributed to environmental causes. In this way, some "functional" mental illnesses have been reclassified as "organic" in nature.

Current Biological Approaches to Mental Illness

According to the dopamine hypothesis of schizophrenia , the symptoms of schizophrenia, and their underlying psychopathology, are caused by excess activity of dopamine, a neurotransmitter substance. Evidence for the dopamine hypothesis includes:

In fact, autopsy studies, and also some brain imaging studies, indicate that there are increased levels of dopamine metabolites in schizophrenic patients.  And that's consistent with the dopamine hypothesis.  Moreover, phenothiazine drugs, which are used in the treatment of schizophrenia, block the neural receptors for dopamine, impairing the uptake of dopamine by post-synaptic neurons.  So the fact that there are increased levels of dopamine metabolites found in schizophrenic patients, and that antipsychotic medications operate to reduce dopamine levels -- these are both factors that are consistent with the dopamine hypothesis of schizophrenia.

But another piece of evidence comes from a laboratory model of schizophrenia known as amphetamine psychosis.  Solomon Snyder and his colleagues, working at the National Institute of Mental Health, found that the administration of certain drugs known as amphetamines can actually produce some of the symptoms of psychosis, particularly schizophrenia, in rats and other laboratory animals.  They also noticed that the habitual heavy use of amphetamines by humans can produce some of the symptoms of schizophrenia as well -- particularly hallucinations, thought disorder, and delusions.  These amphetamine drugs, such as Benzedrine (amphetamine), Dexedrine (dextroamphetamine), and methedrine (methamphetamine) may cause the user to experience hallucinations, while habitual, heavy use can cause thought disorder and paranoid symptoms as well.have the effect of increasing dopamine activity in the brain.  So this laboratory model -- first studied in rats, then in monkeys, and then in humans who abuse amphetamine recreationally -- provides further support for the dopamine hypothesis of schizophrenia.In these ways, amphetamine psychosis seems to mimic at least some of the symptoms usually associated with schizophrenia.

According to the monoamine hypothesis of depression, the symptoms of depression, and their underlying psychopathology, are caused by lowered levels of another class of neurotransmitters, the monoamines, which include norepinephrine and serotonin. Evidence for the monoamine hypothesis includes:

It should be understood that these are only hypotheses about the underlying biology of these syndromes, and that these hypotheses are surely incomplete.

And sometimes biological hypotheses are just wrong. In 1998, a paper published in Lancet by Andrew Wakefield, a British physician, suggested a link between autism and the childhood vaccine for measles, mumps, and rubella (MMR). The resulting concern led to a worldwide reduction in MMR vaccination, as parents hoped to prevent their children from getting autism. The methodology of the study was subsequently criticized, and 10 of Wakefield's 12 co-authors subsequently retracted the paper -- as did the journal in which it was originally published, and Wakefield lost his license to practice medicine in Britain (he relocated to the US). In 2011, an investigation published in the British Medical Journal asserted that Wakefield's paper was not just methodologically weak but actually fraudulent.In fact, subsequent, better designed studies show that there is no evidence for a role of MMR or any other childhood vaccination in autism. Nevertheless, as of 2011, Wakefield (who has relocated to the US) continues to assert such a link, and worldwide MMR vaccination rates have never returned to their pre-1998 levels -- increasing the risk of childhood diseases that are entirely preventable.

For an excellent introduction to the roe of neurotransmitters in mental illness, and the basics of psychopharmacology, see Drugs and the Brain (Rev. Ed., 1996) by Solomon Snyder, one of the deans of psychopharmacology.  Because the pharmaceutical industry is constantly developing new products, the information on specific drugs is necessarily a little dated.  But the basic ideas haven't changed much, and neither has the underlying neuroscience.  It's a good place to start..

Etiology of Mental Illness

Somatogenic and psychogenic theories of mental illness are, first and foremost, theories about the role of nature and nurture, and we have now learned that the proper formulation is nature-nurture questions is not "Which is right?" but rather "How do nature and nurture interact?". The etiology of mental illness is no exception.

Genetics of Mental Illness

One place to look for the origins of psychopathology is in the genes: perhaps certain forms of mental illness, or at least risk factors for them, are passed through families through genetic inheritance.  We know that, for many diagnoses, having a family member with mental illness increases the risk for mental illness in other family members.  Of course, this effect could be environmental as well as genetic.

The study of the genetic basis of mental illness began long before Mendel and research on fruit flies, as the superintendents of 19th-century asylums for the insane or "feeble-minded" traced the family histories of their patients, seeking evidence that mental illness was inherited (and, not incidentally, laying the intellectual foundations for the pseudoscience of eugenics).  For an historical account of this research, see Genetics in the Madhouse: The Unknown History of Human Heredity (2018) by Theodore M. Porter. 

The genetic contribution to mental illness can be assessed by means of the twin-study method described in the lectures on Psychological Development: by comparing the similarity of MZ and DZ twins, we can estimate the contributions of genetics, the shared environment, and the nonshared environment.  When it comes to personality characteristics, such as the Big Five personality traits, similarity is measured by means of the correlation coefficient.  In psychiatric genetics, similarity is more commonly measured by means of the concordance rate -- that is, the probability that two twins will have the same psychiatric disorder.  The calculations for heritability differ a little, but the underlying logic is the same:

This table reflects our best understanding of the heritability of major forms of mental disorder, as of 2012.  The heritability coefficients vary widely, from a low of .37 for Major Depressive Disorder to a high of .80 for Autism Spectrum Disorder and .81 for Schizophrenia.  Note, however, two points about this table:

So, as is generally the case, the origins of mental illness are not to be found in the genotype alone.  Rather, they will have to be found in gene-by-environment interactions (GxE) of the sort discussed under the heading of epigenesis in the lectures on Psychological Development. 

Still, once researchers have established a significant level of heritability for a mental illness, it makes sense to start searching for the genes responsible (for a review, see Duncan et al., 2014).  Before the 21st century, this was not really possible.  We didn't know enough about the human genome, and we didn't have the technology.  And even with the mapping of the human genome, and the availability of (relatively) inexpensive technology, the task is daunting:

Until recently, researchers had to propose, on the basis of some theory, what genes to look for.  So, for example, if they were interested in schizophrenia, they might look at genes that are involved in the production and metabolism of the neurotransmitter dopamine; for depression, they might look for genes involved with the production and metabolism of the neurotransmitters norepinephrine or serotonin.  This is known as the candidate gene strategy.  However, advances in technology have permitted researchers to go on "fishing expeditions" in which they cast their nets more widely, over thousands or millions of candidate genes and their variants, in what are known as genome-wide association studies (GWAS).  These studies have begun to yield results -- and, interestingly, they have been turning up evidence of genes involved in mental illness other than those "candidates" hypothesized by various biochemical theories of mental illness!

It's pretty clear that the search for "the gene" that is "for" schizophrenia, or any other form of mental illness, is going to be complicated (Duncan et al., 2014). 

Once we've determined that there is, in fact, a genetic contribution to some mental illness, the next step is to determine what genes are involved.  I say "genes", plural, because it's clear that there's no single gene "for" schizophrenia, like the gene for blue or brown eyes presented in standard elementary accounts of Mendelian genetics.  Rather, much as with intelligence, the genetic basis for schizophrenia is most likely to involve the cumulative effects of many genes -- dozens, perhaps hundreds.

Recent advances in understanding the genetics of schizophrenia offer interesting insights into the genetic contribution to mental illness more generally.  These advances have been made possible by the Human Genome Project, which in 2003 delivered a map of the human genome, indicating the location of each of our roughly 22,500 genes on our 23 chromosome pairs.  Then began the process of determining the function of each of these genes.  Over the years, gene-mapping has become less expensive and time-consuming, so it is now possible to search the genomes of large numbers of individuals for genes that are associated with various illnesses, employing such techniques candidate gene association, common variant association, and copy number variation. 

This isn't a course in genetics, and if you want details of each of these methods, there's an excellent, highly accessible account of this work in "Runs in the Family" by Siddhartha Mukherjee (New Yorker, 03/28/2016), and a more technical survey in "Genome-Scale Neurogenetics Methodology and Meaning" by McCarroll, Feng, and Hyman (Nature Neuroscience, 2014).  This discussion is largely drawn from these sources.

With data available on a very large number of individuals, investigators are able to identify associations between schizophrenia (and other forms of major mental illness, such as bipolar disorder and autism) with various portions of the human genome.  This task has been undertaking by a group known as the Psychiatric Genomics Consortium (PGC).  The "Manhattan plot" at left (so named because it looks like the New York City Skyline), taken from one such study involving more almost 37,000 patients and more than 110,000 controls, sampled from 20 countries, shows which specific locus on each chromosome is significantly associated with schizophrenia (Sekar et al., Nature 2016).  Of course, with 22,500 genes, a number of these associations could appear just by chance, so the investigators employed appropriate statistical corrections.

• As expected, based on the hypothesis of polygenic inheritance, there were lots of such segments -- more than 100, in fact.
  
• Some of these gene loci are known to be associated with certain neurotransmitters, and others were associated with axonal transmission and other activity within individual neurons.  Presumably, particular alleles of these genes predispose the individual to schizophrenia (or bipolar disorder, or autism, etc.) -- and as more of these alleles accumulate in the individual's genome, his or her risk for major mental illness increases.     
  

These sorts of findings would be expected if, as most theorists believe, major mental illness is fundamentally a disorder of the central nervous system. But other findings were more surprising.  For example, the strongest association (the tallest "skyscraper" in the Manhattan plot above) was on Chromosome 6, in a region known as the major histocompatibility complex (MHC), whose genes are linked to the immune system.  But how do we get from the immune system to schizophrenia?  Here's one prominent theory (Sekar et al., Nature, 2016). 

• The immune system protects the body against disease by detecting and attacking various pathogens in the body.  The genes in question control what are known as complement components (CCs), which identify various viruses, bacteria, and the like for destruction by immune cells.  However, CCs also identify synapses for destruction.  But why would anyone want to destroy synapses? 
• Actually, this is a major feature of neural development.  During the course of fetal development, brain activity creates large numbers of synaptic connections, ready for use when the newborn infant encounters the world.  During learning, neurons that fire close together in space and time (remember long term potentiation?) strengthen their synaptic connections; other, unused synaptic connections are lost, in a process known as pruning (as in pruning trees and bushes).  This pruning process, which goes on for about the first 30 years of life, is a normal process of neural development and plasticity.  The same CCs that identify pathogens for destruction also identify synapses for pruning. 
• A gene in the MHC region of Chromosome 6, known as C4, comes in two forms, known as C4A and C4B, and individuals vary in terms of how many C4A and C4B genes they have, and in what combinations.  C4A and C4B generate corresponding proteins.
  
• A particular variant on C4 elevates the risk for schizophrenia from 1% to 1.27%.  Yes, you read that right.  An elevation of 27%, which sounds like a lot, but in absolute terms still a very small risk.  So C4 alone is not by any means the solution to the genetics of schizophrenia.
• About 27% of the schizophrenic patients in the PGC study had this variant on C4.  But so did 22% of the normal controls!
  
• Mice that lack the C4 gene under-prune their synapses. 
  
• And it is known that the brains of schizophrenic patients contain fewer synapses than those of normal controls.
  
• A certain combination of C4 alleles promoting excessive C4A activity,
  
• This overactivity in the CC system identifies too many synapses for pruning.
• Over-pruning of synapses results in the difficulties in attention and thinking that are characteristic of schizophrenia.
• The normal process of synapse-pruning begins at birth and continues for 20-30 years.  Thus, in theory, over-pruning of synapses results in the typical onset of schizophrenia in late adolescence or the patient's 20s and 30s (schizophrenia used to be called dementia praecox, or "early dementia").
  

Understand: I'm not saying that this theory is the neuroimmunological cue to the secret of the origins of schizophrenia.  It might very well be wrong -- just like most past hypotheses concerning the genetic origins of schizophrenia.

Another multinational group of researchers, known as the Brainstorm Consortium performed a GWAS of 25 different neurological and psychiatric disorders, involving more than 250,000 patients and more than 750,000 healthy controls -- thanks mostly to European countries whose national health systems maintain large databases (Anttila et al. Science, 2018; see also Gandal et al., Science, 2018).  As had proved to be the case in previous, smaller-scale studies, the psychiatric syndromes showed substantial genetic overlap (the only exception was post-traumatic stress disorder, which by definition has a substantial environmental cause).  That is to say, a number of genetic markers were identified for each syndrome, but the same makers tended to appear from one syndrome to another.  By contrast, the 15 neurological syndromes studied, including Alzheimer's disease and Parkinson's disease, showed much more distinct genetic profiles. And there was little overlap between the psychiatric conditions and the neurological ones.  What to make of this isn't clear.  One possibility is that there is a genetic component to risk for mental illness in general, and other factors, whether biological or environmental, determine which specific mental illness a person will suffer.  Another possibility is that the GWAS method, and genetics in general, isn't going to tell us much about the etiology of mental illness.

At the same time, other researchers have turned their attention to the environment, and in particular to those features of the environment that interact with one's genetic heritage.

For more on the search for the genetic basis of schizophrenia, including a plea that researchers spend more time searching for environmental factors that might interact with genetic predispositions, see "Schizophrenia's Unyielding Mysteries" by Michael Balter, Scientific American, 05/2017.

The Diathesis-Stress Model

The origins psychopathology (etiology) may be viewed within the framework of the diathesis-stress model of psychopathology proposed initially by Meehl (1962) and Rosenthal (1963), and elaborated more recently by Monroe and Simons (1991) and Belsky and Pleuss (2009). According to the model:

The diathesis-stress model of psychopathology is a special case of the person-by-situation interaction, where diathesis is an attribute of the person and stress is an attribute of the environment.

In principle, diathesis and stress factors could combine in a number of ways.

In an additive model, diathesis and stress are independent of each other, and the likelihood of an acute episode is simply a function of the sum of diathesis and stress factors. Following Lewin, we might symbolize this situation as E =f(D + S).

In a multiplicative model, diathesis and stress truly interact, so that the combination is truly potent: following Lewin, it would be expressed as D =f(D x S).

Note that diathesis factors are specific to particular forms of mental illness. In theory, some particular diathesis predisposes an individual to schizophrenia, but other specific diathesis would be relevant to depressive disorder, anxiety disorder, etc. In this way, if stress precipitates an acute episode of mental illness, that illness will take a specific form.

Diathesis and Stress in Theory and Practice

Schizophrenia. The diathesis-stress approach was first articulated in the context of schizophrenia. With respect to diathesis, there is a clear genetic component to schizophrenia. Compared to a base rate of about 1% in the population at large, the concordance rate for schizophrenia is clearly elevated among relatives. In monozygotic twins, the concordance rate is about 38%; for dizygotic twins, 8%. Risk for schizophrenia is also increased if a first-degree relative (father, mother, brother, sister) is schizophrenic. Among adoptees, risk is increased if one's biological parent has schizophrenia, but not if one's adoptive parent has schizophrenia. These figures are consistent with the proposition that people can inherit a predisposition to schizophrenia. But note that the concordance rate is far from a perfect 100% (in fact, it appears to be far from even 50%), suggesting that genes are not solely determinative . Any difference between monozygotic twins must be due to the unshared environment, and that is where differences in stress probably come into play. In any event, schizophrenia appears to occur as the product of the combination is of a shared genetic diathesis and an unshared environmental stress.

Rosenthal (1963) originally got his idea about diathesis and stress from the Genain Quadruplets -- identical quadruplet girls, born in 1930 into a family with a history of mental illness, three of whom were hospitalized for schizophrenia on at lest one occasion. "Genain" is a pseudonym, derived from the Greek words for "dire birth", intended to protect the identity of the girls and their family. Because they were studied intensively at the National Institute of Mental Health, they are known as Nora, Iris, Myra, and Hester. What was particularly interesting about the twins was that not all of them fell ill -- at least, Myra was never actually hospitalized. Thus, schizophrenia is not purely a result of genetics -- or else, these identical twins would all have had identical outcomes. Instead, Rosenthal hypothesized that a genetic diathesis interacted with some environmental stress to precipitate schizophrenia in some of the children, but not in all.

Another family challenged by mental illness were the Galvins -- Mimi, Don, and their 12 children, six of whom -- all boys born between 1945 and 1965 -- ended up with a diagnosis of schizophrenia.  The whole sad story is told by Robert Kolker, a journalist, in Hidden Valley Road (2020).  The post-World War II interval was the heyday of psychoanalysis, with its theory that the illness was caused by cold, domineering "schizophrenogenic" mothers who themselves suffered from a "perversion of the maternal instinct".  An entirely bogus theory, like the rest of psychoanalysis, but that didn't prevent Mimi from shouldering the blame; nor did it prevent the boys from getting treatment that might actually have helped them.  Then came the pharmacological revolution, and the boys were so filled with antipsychotic medication that they suffered massive side-effects.  The other six siblings -- four boys and two girls -- escaped the illness, but not really -- because they grew up in an often chaotic household (on Hidden Valley Road) that the parents could barely hold together.  What results, in Kolker's hands, is a compelling depiction of what severe mental illness can do to a family.  And another picture of the complexities of diathesis-stress theory.  Wherever the boys schizophrenia came from, the presence of six mentally ill children in a single family can't help but have piled on the stressors.

One environmental stressor that has been implicated in schizophrenia is socioeconomic status: schizophrenia is rare, affecting less than 1% of the population, but it is more likely to be observed in individuals with relatively low socioeconomic status. One theory is that the stresses of lower-class living interact with a genetic diathesis for schizophrenia, resulting in the higher incidence -- an idea known as sociogenesis. However, careful epidemiological studies have shown that low SES follows, rather than precedes, the onset of schizophrenia. That is, schizophrenia occurs in all socioeconomic strata, but when it happens to upper-class individuals, they tend to drift down to lower socioeconomic strata -- a phenomenon known as social drift.

However, the failure of the sociogenic hypothesis does not rule out environmental contributions to schizophrenia. Other environmental influences that have been causally linked to schizophrenia include:

One environmental stressor that has received a great deal of research attention is deviant communication: Vague and fragmented verbal exchanges, especially on the part of family members. A longitudinal research project known as the Finnish Adoptive Family Study of Schizophrenia examined the long-term outcome of "high risk" children who were born to 167 women hospitalized for schizophrenia and control children born 202 women hospitalized for other illnesses. On medical advice, the children of these women were given up for adoption. In one study, Wahlberg, Wynn, and their colleagues tested the adoptive families for signs of communication deviance, and then tested the adopted children themselves (known in medical terminology as "probands", because their family history gives them an elevated probability of becoming ill) on an index of thought disorder characteristic of schizophrenia.

This is exactly the kind of person-environment interaction anticipated by the diathesis-stress model: the combination of high genetic risk (being the child of a schizophrenic mother)and high environmental stress (being exposed to communication deviance in one's adoptive family) leads to increased occurrence of schizophrenic symptoms. No such trend occurred, however, in children who were not already at risk for schizophrenia.

Now look carefully at the graph. Note that the level of communication deviance in the adoptive families of control children "maxes out" at 8 units, while the level observed in the adoptive families of the high-risk probands goes as high as 10 units. Perhaps, by some stroke of bad luck, these high-risk probands were adopted into families who were carrying more than their fair share of schizophrenic diathesis. More likely, the high-risk probands contributed to the high levels of communication deviance observed in their adoptive families -- communication deviance that might not have been present, but for the probands themselves. Perhaps this is another instance of the person creating the environment to which s/he responds.

Mood Disorder. Studies of bipolar and (especially) unipolar affective disorder show the same patterns: clear evidence for a genetic diathesis, but equally clear evidence for an unshared environmental stress (we cannot calculate the contributions of genetics, shared environment, and nonshared environment from concordance rates in exactly the same way you can for twin correlations, but the logic is the same).

Ulcers. The diathesis-stress approach is also relevant to psychosomatic ulcers, where lesions in the lining of the gastrointestinal system (that's what peptic ulcers are) occur to people who are under a high amount of stress.

The psychosomatic nature of peptic ulcers has recently been discounted by some physicians, who note the presence of a particular bacterial infection,helicobacter pylori, in the stomachs and small intestines of as many as 80% of ulcer patients. But while almost everyone who suffers from ulcers is infected with h. pylori, not everyone infected with h. pylori has ulcers. In fact,h. pylori is also found in the gastrointestinal systems of 70% of patients who do not have ulcers! What makes the difference? Plausibly, stress. Part of the stress response is to secrete acid into the stomach (to aid in the digestion process; when it doesn't encounter food, it eats away at the stomach lining, creating ulcers (gastric ulcers are lesions in the stomach; duodenal ulcers are lesions in the small intestine). Infection by h. pylori increases the gastrointestinal system's vulnerability to ulceration: it is a genuine diathesis factor. But, prolonged stress-related autonomic activation (see the discussion of Selye's general activation syndrome in the lecture supplements on the Biological Basis of Mind and Behavior) can interact with the bacteria to make ulcers even more likely to occur.

The diathesis-stress hypothesis is supported by a laboratory model of ulcers developed by Steven Maier (one of the investigators involved in the discovery of learned helplessness). In this model, rats are infected with h. pylori bacteria, and then are exposed to stress in the form of unpredictable and uncontrollable shock. This combination is particularly likely to produce ulcers, compared to conditions in which neither diathesis nor stress, or only one factor but not the other, are present.

Phobias. In experimental psychopathology, phobias are a classical example of psychopathology acquired through learning -- particularly, fear conditioning. As such, phobias would seem to be a case of all stress and no diathesis: the stress is the anxiety that accompanies exposure to the feared object. So, if a person has a negative encounter with a snake, he or she will come to fear snakes. In this conditioning theory of phobia, the snake is a CS that predicts unpleasant consequences.

This is a fine theory, so far as it goes, but it has two problems.

One problem is that people with phobias don't always, or even usually, have histories of negative experiences with the objects of their fears. Readers who have phobias concerning snakes, for example, might ask themselves what snakes have ever done to them. Once in a while a snake phobic has been bitten by a snake, but not too often. Instead of resulting from direct experience with the phobic object, it is more likely that the snake phobia has been acquired through social learning or vicarious conditioning. That is, people become afraid of snakes because they know other people who are afraid of snakes. We learn to fear what other people fear, without having frightening experiences ourselves.

The second problem is that people don't always acquire phobias following association of an object with negative consequences. To use an example from Seligman (the same theorist who proposed the learned helplessness model of depression), when we have a bout of food poisoning we don't become afraid of the crockery and cutlery; we become afraid of the food. And not just any food we may have eaten; we tend to become afraid of thinks like Lima beans and cream sauces. In fact, clinical phobias are largely limited to a relatively small number of situations: open spaces, high places, the gaze of other people, and wriggly, slimy things. According to Seligman, we are prepared by evolution to easily and quickly acquire conditioned fear responses to these sorts of objects and situations. In this view, the diathesis in phobia is a set of "prepared" associations, a part of the organism's evolutionary heritage, which predispose the individual to acquire intense fears even with minimal exposure.And the stress is a negative event. The stressful event can result in phobic levels of fear, but only by virtue of these prepared associations.

A laboratory model of phobias incorporating both social learning and preparedness has been studied by Mineka and her colleagues in research described in the lecture supplement on Learning. Mineka and her colleagues showed that observational learning was sufficient to produce intense conditioned fear in monkeys who themselves had no experience of negative consequences in association with the CS: they learned to fear what other monkeys feared. But Mineka et al. also found that observational learning didn't produce fear of just anything. Through vicarious learning, monkeys acquired conditioned fear responses to snakes, but not flowers. According to the preparedness argument, a disposition to fear snakes is built into monkeys by evolution, and can produce full-blown snake-fear even in with little or no direct experience.

The Dunedin Studies. The interaction of a biological diathesis with environmental stressors can be illustrated by two studies by Avshalom Caspi, Terrie Moffitt, and their associates, based on data collected in the "Dunedin Multidisciplinary Health and Development Study". In this project, longitudinal data was collected from a "birth cohort" of 1,037 children (roughly half of them males) born near Dunedin, New Zealand, and tested approximately every two or three years from ages 3 to 26 (actually, they're still being followed).

In one study, Caspi et al. (2002) examined the role that the MAOA gene played in adolescent conduct disorder. This gene, located on the X chromosome, promotes monoamine oxidase A, a substance that metabolizes many different neurotransmitters, and which has been linked to increased aggression in both laboratory mice and humans. Caspi et al. also explored the role of stress in conduct disorder -- particularly a history of childhood and adolescent maltreatment, which some theorists have proposed initiates an intergenerational "vicious cycle of violence" in which maltreated boys become maltreating fathers, producing maltreated boys who also become maltreating fathers. In fact, subjects with high levels of MAOA activity showed a relatively low incidence of conduct disorder, regardless of their history of maltreatment. However, subjects with low levels of MAOA activity, who also had a history of severe maltreatment, showed a very high incidence of conduct disorder. The MAOA gene is a diathesis which interacts with severe maltreatment to produce conduct disorder.

In another study, Caspi et al. (2003) examined the role of the 5-HTT gene in major depressive disorder. This gene, located on chromosome 17q11.2, comes in two forms, "short" and "long", yielding four genotypes: SS, SL, LS, or LL. Caspi et al. also explored the role of life stress in depression, by counting the number of stressful events occurring in the life of each subject between ages 21 and 26 (in psychology, a "stressful" event can include getting married as well as getting divorced). Subjects with the LL form of the genotype showed a relatively low incidence of depression, regardless of their history of life stress. However, subjects with the "short" form of the genotype (with at least one short allele, as in SS, SL, or LS), combined with a history of many stressful events during the previous five years, showed a much higher incidence of depression. 

Similarly, Fox et al. studied the role of the 5-HTT gene in pathological shyness -- children and adults who are severely withdrawn. Their study actually involved multiple assessments of children's temperament -- how inhibited the children were in the presence of strangers, and their mothers' ratings of their shyness. The mothers also provided ratings of the amount of social support (e.g., friends) their children had. The results of the study showed a clear gene x environment interaction:

Notice the shape of the graphs in the Fox et al. study of pathological shyness, which combines the "crossover" and "fan" effects that are so characteristic of the person-by-situation interaction.

Still another study from the Caspi-Moffitt group focused on marijuana and psychosis.  It’s long been known that some people who smoked marijuana as adolescents develop a form of psychosis as adults, but the precise pathway has been unclear.  Certainly, most adolescents who smoke marijuana don’t develop psychosis, but it does appear to be a risk factor.  Caspi and Moffitt focused their attention on yet another gene, known as COMT, located on Chromosome 22. COMT is involved in the metabolism of dopamine, which has been linked to schizophrenia.  The gene comes in two forms, methionine (“Met”), and valine (“Val”).  Individuals who have two copies of the “Met” allele show the fastest breakdown of dopamine; those with two copies of “Val”, the slowest; and those with one of each, somewhere in between.  So, if you’re a “MetMet” person, dopamine metabolizes faster, and resides in your system for less time.  Again using subjects from the Dunedin study, Caspi and Moffitt classified their subjects according to the form of the COMT gene, and also by their history of adolescent marijuana use.  And when they looked at the incidence of psychotic symptoms in these subjects when they were young adults, they found a clear gene-by-environment interaction.  The affected subjects didn’t always show full-blown schizophrenia or any other psychotic syndrome.  Still, their risk for delusions, hallucinations, and other “schizophreniform” symptoms was greatly increased if they had two copies of the “Val” allele, coupled with frequent marijuana use as adolescents.  If they had only one copy of “Val", or two copies of “Met”, their risk was greatly reduced.

Another study showed how the COMT genotype interacted with stress to affect performance on a academic tests (Yeh et al. 2009) -- not, admittedly, a major mental illness, but perhaps indicative of GxE interactions in anxiety disorders.  Every year, Taiwanese students who wish to move on from junior to senior high school must take a rigorous test known as the Basic Competence Test, which measures educational achievement in a number of subjects, including Chinese and English language, mathematics, science, social science, and writing.  The BCT is an extremely stressful "high-stakes" test, because its outcome determines whether the student will have any chance of going to college (at least in Taiwan).  Yeh drew a sample of 779 Taiwanese high-school students (i.e., who had already passed the BCT), and examined their scores on the BCT subtests as a function of their COMT genotype.

The conclusion is that having two copies of the Met genotype renders the person vulnerable to high levels of stress, to the detriment of performance.

Some of these gene-environment interactions are controversial. 

In particular, the gene-by-environment interaction in depression, involving the 5-HTT gene, has stimulated a great deal of interest, but it’s also been controversial.  Some researchers have failed to replicate Caspi and Moffitt’s findings, while some critics have complained about the assessment of stress.  Katja Karg and her colleagues recently surveyed 56 studies, involving more than 40,000 subjects, and found that, overall, these studies confirmed the G-by-E effect.  A history of stress, especially defined by childhood maltreatment or life-threatening or chronic medical conditions, coupled with the “short” form of the 5-HTT gene, greatly increases one’s risk for a major depressive episode. So please note that the findings on 5-HTT or COMT, for example, are not to be taken as firm.  Rather, I cite them here simply as illustrations of a particular approach, using the diathesis-stress model, that is commonly used to identify the genetic and environmental contributions to mental illness.  So stay tuned!

The Nature of Diathesis and Stress

In standard presentations of the diathesis-stress model, the diathesis is often biological (like the 5-HTT gene) and the stress is often psychosocial (like stressful life events), but this need not necessarily be the case.

Usually, we think of biological diatheses as specific genes, like MAO-A or 5-HTT.  But it's possible that there is a genetic diathesis for a broader group of mental illnesses.  Smoller and his colleagues (2013) conducted the largest study to date of the genetics of mental illness, including more than 60,000 subjects from 19 countries (roughly half were patients carrying a psychiatric diagnosis).  They found that five different disorders -- schizophrenia, bipolar disorder, depression, attention deficit hyperactivity disorder, and autism -- shared a relatively small set of genetic aberrations in common.  For example, one identical twin might develop schizophrenia, while another might develop bipolar disorder. 

Somewhat similar results were obtained from another study which extracted RNA (not DNA) from the cerebral cortex of deceased patients who had been diagnosed with various forms of major mental illness, such as autism, schizophrenia, bipolar disorder, depression, and alcoholism (Gandal et al., Science, 2018).  They compared these assays to samples taken from individuals who did not carry a psychiatric diagnosis (an obvious control), and from other patients with a physical illness, irritable bowel syndrome (in order to control for illness in general).  As in the Smoller study, they found that there was significant overlap in the patterns of gene expression across the various syndromes -- except alcoholism, which had a pattern that was quite distinct compared to the others.  For example, there was considerable overlap in gene activity between schizophrenia and bipolar disorder -- despite the fact that the symptoms associated with these two syndromes are very different.

These studies suggest that there might be a genetic diathesis for these major forms of mental illness in general, and that whatever specific mental illness an affected individual develops would depend on other factors (perhaps environmental, perhaps genetic as well). 

Biological Stressors

In some instances, stress is better conceptualized as biological rather than psychosocial in nature.

For example, prenatal and perinatal complications are often found in the life histories of people who eventually develop schizophrenia. These are environmental stressors, but the fact that they occur in the prenatal or perinatal environment mark them as more biological than psychosocial in nature. A difficult birth doesn't have the same meaning for a person as a difficult childhood or adolescence.

Psychosocial Diatheses

In other instances, diathesis is better conceptualized as psychosocial rather than biological in nature

For example, people may be predisposed to depression by histories of social learning that lead them to acquire certain beliefs.

A similar account could be given of the depression which occurs in some (but not all) women who are going through menopause.

Post-traumatic stress disorder (PTSD) would, at first glance, appear to be an almost "pure" case of stress-related mental illness. But, in fact, only a minority of people actually exposed to traumatic levels of stress actually go on to develop PTSD (there is, of course, an acute stress disorder affecting large proportions of trauma victims, but even this isn't universal).  This variability was noted even as far back as World War I, where it was attributed to individual differences in soldiers' predisposition to stress.  This diathesis might be biological -- perhaps something related to the functioning of the hypothalamic-pituitary-adrenal (HPA) axis, or perhaps a genetic predisposition.  Or it might be something psychological, analogous to the depressogenic schemata and attributional styles implicated in some forms of depression.  But the basic point is that even in PTSD, the stressor alone is rarely sufficient to cause the disorder all by itself; the stressor has to combine with a predisposition or vulnerability.
Interestingly, the role of diathesis or predisposition in PTSD has contributed to the reluctance to award honors or disability benefits to some soldiers suffering from PTSD.  The argument is that the soldiers' disabilities weren't caused by the stress of war, but rather are related to personality problems that predated the war -- much as insurance companies will sometimes deny medical coverage on grounds of a "pre-existing condition".  Setting this policy issue aside, PTSD may be an example where both the diathesis and the stress are psychosocial in nature.

Diathesis is often biological, and stress is often psychosocial, in nature. But the really important feature of diathesis is that it is something that the person carries with him into a situation, either as a biological or a psychosocial "trait". By the same token, the really important feature of stress is that it is something that happens to the person in a particular situation, ether as a biological or psychosocial event (or series of events).

Treatment of Mental Illness

The diathesis-stress model for the origins of mental illness offers a framework for understanding their treatment and prevention as well. So, if mental illness is caused by the interaction of diathesis and stress, then effective interventions should alter diathesis factors, stress factors, or both.

To orient ourselves, let's first examine what happens during an episode of mental illness.  As in earlier lectures in this module, we'll be using language derived from medicine patient, symptom, syndrome, and so on, to identify what we can call, following Steven Hollon, a prominent depression researcher at Vanderbilt University, and others, the 5 Rs of Mental Illness. I'll use depression as my example, but the same general points would apply to any syndrome of mental illness.

First, let's assume that the patient begins at a more-or-less "normal" state, with no identifiable symptoms of mental illness.  At some point, however, given a certain combination of diathesis and stress, he or she begins to show symptoms of some form of mental illness -- depression, or anxiety disorder, or schizophrenia.  At some point, enough symptoms develop, with enough severity, to compromise normal functioning, at which point we can say that the person has "progressed" -- that's the medical term -- to a full-blown acute episode of mental illness.

1. Remission.  With time, even without active intervention, the symptoms may disappear on their own, in which case we would talk about spontaneous remission of the illness.  "Remission".  Spontaneous remission is surprisingly common in depression, which tends to come and go in its natural course.  But it also takes time, maybe 6-9 months, so rather than wait the illness out, it's probably better to get some sort of active treatment.
   
2. Response.  Now let's suppose that the patient does receive some active treatment, whether it's psychotherapy, or some kind of biological treatment, like medication.  Many patients will respond positively to whatever treatment they receive.  If they don't, the therapist may switch to a different form of treatment -- a different approach to therapy or perhaps a different drug -- until something seems to work.
   
3. Recovery.  In depression, subjects may respond positively after a couple of weeks of appropriate treatment, but it may take a while before they really get back to normal -- a complete recovery, in which they are symptom free.  Alternatively, may show a partial recovery, in which some symptoms remit completely, but not others; or there may be a clinically significant reduction in the severity of symptoms.  Sometimes, unfortunately, recovery doesn't occur at all, in which case the patient proceeds from an acute to a chronic illness, calling for continuation treatment.
   
4. Relapse.  Even with a complete recovery, there is some chance that the symptoms may come back, perhaps not as severely as before, but enough to interfere with normal functioning.  In depression, for example, the likelihood of a relapse even after what appeared to be a complete recovery is about nine times the risk of depression occurring in the general population.  Apparently the acute episode is still going on, appearances to the contrary notwithstanding.  For this reason, most therapists don't stop treatment immediately once the patient has recovered, but continue it for a while. 
   
5. Recurrence.  Even after a patient has completely recovered, there is still some chance that another acute episode can occur later.  If a person has recovered from an acute episode of depression, there is still a chance that another acute episode will occur -- about three times more risk than in the general population.  For this reason, even patients who have completely recovered may want to maintain some degree of maintenance treatment. 
   

The first goal of treatment is to achieve a cure. Genuine cures will do more than suppress superficial symptoms: they will eliminate underlying pathology, or reduce it to a clinically significant extent, returning the person's level of functioning to "within normal limits" even after active treatment has stopped. In the absence of a cure, treatment focuses on the amelioration of symptoms, or rehabilitation regimes that permit the patient to cope with a chronic condition.

Custodial Care

Historically, the scientific treatment of mental illness has come in three basic forms.  Up until the 20th century, there were really no active treatments available for mental illness, so intervention focused largely on custodial care -- chiefly the warehousing of the mentally ill in public hospitals or private asylums.  Pennsylvania Hospital, America’s first hospital, was founded by Benjamin Franklin and Thomas Bond in 1751, when Pennsylvania was still a British colony, for the care of the indigent and the mentally ill.  Initially, the asylum was simply a separate ward in the original hospital on Pine Street, later a separate wing, the asylum was moved to the more rural West Philadelphia in 1841, and in 1859 to its ultimate location (at 49th and Market Streets). 

• Originally, the asylum housed only dangerous and disruptive "lunatics", and the primary "treatment" was restraint, along with purging and bloodletting (the humor theory of disease was still popular, and it applied to mental as well as medical illnesses). 
  
• Benjamin Rush, a physician at the Pennsylvania Hospital from 1783 to 1813 who was among the signers of the Declaration of Independence, took such an interest in the treatment of the mentally ill that he is now considered the "father" of American psychiatry.  In many respects, Rush was a conventional physician for his time, when the "standard of care" emphasized bleeding and purging.  On the other hand, he was the first American physician to apply what we now call the "medical" model, as opposed to the "moral" or "supernatural" models, to mental illness.  His view that mental illnesses were true "diseases of the mind" which should be freed from moral stigma, and be treated with medicine rather than moralizing, marks a shift from the moral model of mental illness to the medical model.  Influenced by the Quakers who established an asylum in the nearby Friends Hospital, Rush pioneered a set of reforms for the treatment of the mentally ill -- just as he had earlier reformed Pennsylvania's penal system at the Eastern State Penitentiary.  For example, he established "day rooms" where patients could be visited by family and friends, and have access to books, game, and the like.  Rush's first son, John, apparently suffered from depression.  After attempting suicide in New Orleans in 1810, he was brought home to Philadelphia and lived at the Pennsylvania Hospital until he died in 1837 (the elder Rush himself died in 1813).
• For recent biographies of Rush, see Rush: Revolution, Madness, and the Visionary Doctor Who Became a Founding Father (2018) by Stephen Fried and Dr.  Benjamin Rush: The Founding Father Who Healed a Wounded Nation (2018) by Harlow Giles Unger.
  
• Another set of reforms was instituted by Thomas Story Kirkbride, who was the founding superintendent of the new asylum in West Philadelphia, then known as the Pennsylvania Hospital for the Insane and later at The Institute of Pennsylvania Hospital.  Kirkbride had been a medical resident at Friends Hospital, and was impressed by the "moral treatment" of the patients there.  Like Pinel before him, Kirkbride freed even "dangerous" patents from their restraints, and permitted to walk around the wards outside their hospital rooms.  Patients on a ward took their meals together, instead of in their separate rooms, and were encouraged to read, play games, or work on the hospital grounds.  In 1844, Kirkbride was one of the founders of the Association of Medical Superintendents of American Institutions for the Insane, which eventually became the American Psychiatric Association.  In 1854, Kirkbride published his treatise On the Construction, Organization, and General Arrangement of Hospitals for the Insane, which set the standard for state and private mental hospitals for many decades -- so much so that the hospitals of this era are often called "Kirkbrides".  Kirkbride got to put his ideas into practice when the new hospital was built in 1859.  The hospital was deliberately placed far from the center of town (which was true then), on a piece of land so large as to contain gardens and even a small farm, offering patients the opportunity to work and play outside.  There was a wall for security, but it was placed so far from the hospital buildings as to be unobtrusive.  The buildings themselves were completely up-to-date in terms of plumbing, heating, ventilation, lighting, and the like.  There were two wings, one for each sex, with the most severely ill patients placed on the ground floor farthest from the central administrative offices, and the least impaired on upper floors close to the core.  The corridors were wide, the ceilings high, there were lots of windows, and lots of doors to the outside.  There was a library and a game room.  Many of the staff lived on the hospital grounds, including the superintendent himself. 
  
• The Kirkbride model was followed by at least 75 other private and public mental hospitals in the latter portion of the 19th century, including the Oregon State Hospital, which was the setting for the film of Ken Keesey's novel, One Flew Over the Cuckoo's Nest. 
  
• Another "Kirkbride", the Buffalo State Asylum for the Insane (designed by the great 19th century architect H.H. Richardson), is now the Hotel Henry.  St. Elizabeth's Hospital, in Washington, D.C., is slated to become the new headquarters of the Department of Homeland Security (no wisecracks!).
  I worked at the Institute as a graduate student in the early 1970s, and it seemed as much a resort as a hospital. 

New York Hospital and the Virginia Asylum were also chartered for the care of “Lunaticks” before the Revolutionary War, but did not open their doors until after the War was over.  Still, the patients there were mostly “warehoused”, out of sight of the rest of the community.  After the Civil War, Silas Weir Mitchell gave a more positive spin to “warehousing” by prescribing a “rest cure” for “mental exhaustion – famously depicted in The Yellow Wallpaper, an early feminist classic by Charlotte Perkins Gilman. 

Beginning in the late 19th century, various forms of psychotherapy were introduced.  These interventions, one way or another, were intended to alter the patients' pathological mental states -- their abnormal beliefs and feelings and desires, which were thought to underlie the various symptoms of mental illness.  By changing these abnormal, maladaptive mental states, mental health professionals sought to change the maladaptive patterns of behavior that brought patients to the attention to the professionals in the first place.  Beginning in the 20th century, a wide variety of biological treatments were introduced for mental illness.  These began with procedures like electroconvulsive therapy, and even psycho-surgery where physicians operated on various brain centers that were presumed to be implicated in the patient's problems.  But more recently, a wide variety of medications have been introduced for the treatment of various metal illnesses, beginning with schizophrenia.  These medications have largely supplanted psychosurgery and ECT, though both still have a place in selected cases. 

Biological Treatments

Where biological causes are the primary factor in mental illness, strictly biological treatments may effect a complete cure. Of course, such biological cures require an understanding of the biological basis of some form of mental illness.

Biological cures are exemplified by treatments available for two specific forms of intellectual disability.

There are many other biological treatments for mental illness, including many different medications but also psychosurgery and electroconvulsive therapy. These are often very effective, but they do not reach the level of a cure, because the illness returns when the treatment is discontinued.

Psychosurgery, especially prefrontal lobotomy (the destruction of the prefrontal cortex of the brain) was once rather popular: Egas Moniz, the Portuguese neurologist who pioneered the technique of "prefrontal leukotomy", even won the Nobel Prize for Physiology or Medicine in 1949. The technique has now been completely repudiated, and Moniz' prize considered something of an embarrassment.  However, as our understanding of brain function advances more nuanced surgical approaches to mental illness are sometimes proposed. For a history of psychosurgery, see Great and Desperate Cures: The Rise and Decline of Psychosurgery and Other Radical Treatments for Mental Illness (1986) by Elliot Valenstein, a distinguished neuroscientist. Valenstein's other books are also of interest:Brain Control: A Critical Examination of Brain Stimulation and Psychosurgery (1977) and Blaming the Brain : The Truth About Drugs and Mental Health (1998).

Although nobody does lobotomies anymore, psychosurgeries are still performed (of course, brain surgery is frequently performed for strictly neurological disorders like epilepsy).  Most of these operations occur in cases of obsessive-compulsive disorder or depression, and they are pretty rare -- although, with increasing evidence of efficacy and safety, their use is increasing (though they are not risk-free).  Most of these "surgeries" are performed by means of electrical stimulation through microelectrodes, as opposed to with a scalpel (as was the case, for example, with the earlier lobotomies).     

The treatments are strictly experimental, and they can have powerful and unpleasant side-effects, so they are typically performed only under a "humanitarian device exemption", where no other treatment has worked.  For this reason, deep brain stimulation has not been subject to the kinds of controlled clinical trials that are required for approval of new medications and other medical treatments.

Electroconvulsive Therapy (ECT) is sometimes used as a treatment for severe depression. ECT employs a brief electrical current applied to the scalp to induce a brief seizure somewhat similar to that seen in epilepsy. ECT has a somewhat unsavory history, having been misused in the past, but the fact is that a short course of ECT, judiciously applied, can often produce rapid remission of depressive symptoms. ECT is often avoided because of its presumed side-effects, which (when misapplied) can include brain damage. However, ECT does not hurt (because the patient is unconscious during the treatment). ECT applied bilaterally (with electrodes placed on both sides of the head) can produce an amnesia similar to that which occurs following a concussive blow to the head; but memory impairment can be reduced substantially if the treatment is applied unilaterally, so the seizure is confined to one (usually the non-dominant) hemisphere. Biologically, ECT increases levels of norepinephrine and serotonin in the brain, which is interesting in light of the "monoamine" hypothesis of depression. 

By far the most popular and effective biological treatments for mental illness involve psychotropic or psychoactive medications: beginning in the 1950s, psychiatry has been a full participant in the "pharmaceutical revolution" in medicine.

A variety of antipsychotic medications are used in the treatment of schizophrenia. Most common, perhaps, are the phenothiazines, which appear to decrease dopamine levels in the brain. Typical brand names are Thorazine, Stelazine, Prolixin, and Mellaril. Haldol (a butyrophenone), and Navane (a thioxanthene) are also widely used. There is also a group of "atypical" antipsychotics like Clozaril, Resperidal, Zyprexa, and Abilify, which are as effective as the earlier antipsychotic agents, but cause fewer side-effects.

To be honest, though, most of the drugs used in the treatment of schizophrenia are really nothing more than major tranquilizers, which calm the patient, reducing the tendency to report and act on hallucinations if not to have them.Setting aside the dopamine hypothesis of schizophrenia, nobody thinks that these drugs act directly on the patient's underlying pathology. However, these drugs do make patients more manageable, and their introduction made everybody's life better, staff and patients, inside mental hospitals -- and made it possible for many people with schizophrenia to be released from mental hospitals to live with their families or elsewhere in the community. But, as with many psychiatric drugs, you've got to keep taking them. And once released from direct supervision, this is something that many schizophrenics just aren't inclined to do.

A major revolution in the treatment of depression came with the introduction of antidepressant drugs.

The SSRIs and SNRIs are as effective as the earlier tricyclics and MAO inhibitors -- a fact that has led to the revision of the general monoamine hypothesis of depression by the more specific serotonin hypothesis of depression.

For a discussion of recent advances in the drug treatment of depression, see

• Lifting the Black Cloud" by R.M. Henig,Scientific American, 03/2012.

Also within the category of mood disorders,lithium carbonate has proved to be a very effective treatment for bipolar disorder, also known as "manic-depressive illness". Treatment with lithium reduces or eliminates episodes of mania in as many as 70% of cases, and also reduces episodes of depression. However, because lithium is toxic, its use must be carefully monitored. Lithium is another one of those psychiatric drugs which seem to work -- in fact, it works very well, and works only for bipolar disorder -- but we have no idea why it works, or what bearing its effectiveness might have on our understanding of the nature of mania or bipolar disorder.  Certainly, mania isn't caused by a lack of lithium in the body. 

Just like Coca-Cola once actually contained cocaine, so 7-Up, another popular soft drink once contained lithium.  This just one of the fun facts included in Lithium: A Doctor, a Drug, and a Breakthrough (2019) by Walter Brown, a psychiatrist.  Discovered in 1949 John Cade, an Australian psychiatrist, lithium was the very first psychiatric drug to approach anything like a cure, as opposed to mere symptom relief (remember, before this time, all psychiatric medications were just tranquilizers).  And the research which first demonstrated its curative effects, carried out by Mogens Schou, a Danish psychiatrist, in 1954, were the first randomized, controlled clinical trials of any psychiatric drug.

Similarly, early drug treatment of anxiety disorder focused on sedatives such as the barbiturates (e.g. Nembutol and Seconol), propanediols (e.g., Miltown, Equanil), and benzodiazepines (e.g., Librium, Valium, and Xanax). The 20th century was known as the "Age of Anxiety", and popular literature and movies concerned with American middle-class in the 1950s and 1960s are full of references to these drugs. These drugs are also, essentially, tranquilizers -- though far less potent than those used in the management of schizophrenia. Specifically, they increase the activity of the neurotransmitter GABA, which in turn suppresses activity in the hypothalamic-pituitary-adrenal axis (HPA).

As with the antipsychotic medications, there is also a group of atypical anxiolytics, such as Buspirone, which are as generally effective as the earlier sedatives but with fewer side-effects.

Whether "typical" or "atypical", all psychotropic drugs in a particular class have the same basic pharmacological mechanism of action:

Newer drugs within a class are sometimes called "me-too" drugs, because they all have the same basic effects -- put another way, the newer drugs are little more than copycats of the older ones.  The newer drugs may have fewer side-effects, and be somewhat safer and easier to tolerate.  But the basic underlying pharmacological action is no different.  So why are new drugs introduced?  Mostly because the patents are running out on the old ones.  And, in fact, the pharmaceutical industry isn't devoting that much research to the development of truly new psychotropic drugs, because -- despite their popularity -- there isn't that much to be gained financially from them.  There is much ore money to be made from drugs that treat cancer, heart disease, or diabetes.  

Because of the co-morbidity between anxiety and depression, anxiety disorder is sometimes treated with SSRIs such as Paxil. The rationale for this strategy isn't completely clear -- maybe it's to help patients feel less depressed about being anxious!

In many instances, specific drug treatments are developed based on a specific somatogenic theory of the illness in question. In theory, at least, psychotropic medications work because they address the biological bases of mental illness. Thus, according to the dopamine hypothesis of schizophrenia, drugs that alter the processing of dopamine should be effective treatments for schizophrenia but not depression; and according to the amine hypothesis of depression, drugs that alter the processing of monoamine neurotransmitters should be effective treatments for depression but not for schizophrenia. The SSRIs, which act specifically to enhance the availability of serotonin in depressed patients, are perhaps the clearest expression of this connection between theory and treatment.

But sometimes drug treatments are simply empirical, meaning that they are prescribed simply because they are known to help, and not because their efficacy is predicted by any theory of the illness in question.

A good example of such an empirical treatment is the prescription of amphetamine, such as Ritalin, for the treatment of ADHD.

• A report by Bradley (1937) had noted that Benzedrine, a trade name for amphetamine, had positive, and paradoxical, effects on a group of schoolchildren with various "behavioral issues" -- what we would now recognize as ADHD.  Bradley originally prescribed the drug to treat headaches, and noticed that it affected their behavior as well. 
  
• That same year, interestingly, news media carried reports of college students taking amphetamine as a study aid.  Plus ca change.... 
  
• Ritalin, another proprietary form of amphetamine, was put on the market in 1956, and a study soon showed that it, too, produced positive effects on children with various behavioral disturbances (Ritalin was named after the wife of the chemist who formulated the drug). 
  
• In 1970, amphetamines were placed under tight restrictions (Schedule II) by the Controlled Substances Act.  Initially, sales of Ritalin and other amphetamines declined substantially, but recovered as more children were diagnosed with hyperactivity disorder or ADD/ADHD.
• Another proprietary amphetamine, Adderall, was introduced in 1996.  Essentially the same drug was formerly marketed as Obetrol, for the control of obesity; its new name is a contraction of "ADD for all", which should tell you something.
• Concerta was introduced in 2000.
• The landmark "MTA" study (the acronym stands for "Multimodal Treatment of Children with ADHD") found that 68% of children who received Ritalin or Concerta with behavior therapy showed improvement; comparable figures for medication alone were 56%, behavior therapy alone 34%, and 25% for an untreated control group.  This study effectively set medication as the "standard of care" for ADHD.  
  

There's no good reason why Ritalin should work in these cases -- and given that amphetamine is a central nervous system stimulant, there would be every good reason to think that it would make things worse. Such "off-label" use of medications is a fairly common practice in medicine: physicians may "experimentally" prescribe drugs for conditions other than those indications approved by the Food and Drug Administration following controlled studies of safety and efficacy). But exactly how it occurred to anyone to try Ritalin in the first place isn't clear. Perhaps it was simply an act of desperation after everything else failed. But it does work, at least for many patients with ADHD, and is now the standard of care for both children and adults carrying the diagnosis -- ADHD is now a "labeled" indicator for Ritalin and similar drugs, rather than an "off-label" use. But there's still no theory that explains the paradoxical effects of Ritalin on ADHS. One theory is that amphetamines activate brain centers for the control of attention which are relatively inactive in patients with ADHD -- but this is just a theory, and a rather post-hoc one at that, and as yet there's no evidence supporting it.

Actually, there is a double paradox here, because the evidence for the efficacy of amphetamines in the treatment of ADHD is somewhat ambiguous.  Laboratory tests show that Ritalin (and similar drugs) improves performance on laboratory tests of attention and working, and also has positive effects on brain centers involved in these cognitive functions.  But other studies show that there is no corresponding improvement in student academic performance, as measured by grade-point averages, achievement-test scores, or even the likelihood of repeating a grade. of elementary and secondary-school.  So, whatever is going on in the laboratory, and the brain, doesn't seem to translate into the real life of the classroom.  One possibility is that these children, and their families, come to rely too heavily on the drugs.  In the absence of proper study skills, and a home environment conducive to and encouraging of study, academic performance won't improve just by taking a pill.

The pharmaceutical revolution in mental health has been a genuine revolution, providing a degree of symptom relief that simply was not available previously, and enabling the de-institutionalization of large numbers of patients from mental hospitals, which offered mostly custodial care, back to their homes and into the community; it also improved the lives of a large number of individuals who were being treated on an outpatient basis by psychiatrists and other mental health professionals.

The effectiveness of these drugs also has theoretical implications for our understanding of the biological mechanisms involved in certain forms of mental illness.

In theory, these drugs attack the biological bases of the symptoms of these disorders -- the biological bases of their underlying psychological deficits. But note that the reasoning here is somewhat circular: How do we know that dopamine is implicated in schizophrenia? Because the phenothiazines, which act on dopamine circuits in the brain, are effective in the treatment of schizophrenia. Why do the phenothiazines work so well? Because excess dopamine is the cause of schizophrenia. What we really need is independent evidence that these neurotransmitters are specifically involved in these forms of mental illness.

Getting Past the Blood-Brain Barrier Psychotropic drugs are typically delivered through pills or injection, but either way they have to get to the brain to be effective.  Technically, that's a problem because the brain has a built-in defense against foreign substances, known as the blood-brain barrier (BBB).  The blood vessels in the brain are structured somewhat differently than those found elsewhere in the body.  They are lined with a tightly packed sheath of lipid (endothelial) cells that prevent pretty much anything except oxygen and glucose from getting through to the neurons.  Various pathogens and ions, as well as some proteins that can harm neural cells, are filtered out.  And so are many potentially useful psychotropic drugs -- which is one reason why psychotropic drugs are so hard to develop.  For a thorough discussion of this problem, see "A Barrier to Progress: Getting Drugs to the Brain" by Rachel Brazil, Pharmaceutical Journal, 05/15/2017, from which the image is taken.  Fortunately, there are ways around -- or, more accurately, through the BBB.  Some molecules pass through because they're soluble in water: they can cross the BBB on their own, through a process known as paracellular transport.  This is very rare, because the junctions between epithelial cells is very tight.  Others are small enough that they are soluble in lipids.  Since the BBB is made up of lipid cells, they pass through the epithelial cells themselves -- diffusion.  These drugs include some antidepressants and many other psychotropic drugs, including many addictive drugs such as alcohol, nicotine, morphine, and heroin. Other substances, such as biologics, which involve lager molecules, can be carried through the epithelial cells through specialized protein transporters, such as amino acid chains, in what is sometimes known as the "Trojan Horse approach".  There are a number of variants on this "hitchhiker" strategy, but the pharmacochemistry involved is much more than is required at the level of an introductory psychology course.  This approach is especially popular for delivering chemotherapy in the treatment of brain cancer. Finally,  there are drugs that disrupt the BBB itself, briefly opening a window that allows the molecule through the epithelial cell.  This last technique offers clues to the pathology of Alzheimer's disease (AD) and other dementing illnesses.  The whole point of the BBB is to prevent toxins and other harmful substances from reaching the brain.  If you disrupt the BBB, then you open up the opportunity for such substances to get in.  There is already evidence that the BBB deteriorates with age.  This can allow proteins such as albumin, which ordinarily cannot cross the BBB, to get through, initiating a sequence of events that can result in the brain damage seen in dementia -- like the plaques that are characteristic of AD.  So, one approach to treating, or perhaps preventing, AD is to find a way to "plug" the leaks in the BBB, stopping the cascade to dementia before it can start or gain momentum. For more on the role that a "leaky" BBB might play in AD and other forms of dementia, see "Holes in the Shield" by D. Kaufer and A. Friedman, Scientific American, 05/2021, from which this image is taken.

Setting aside the positive effects of these drugs, it's also the case that pharmacotherapy has some problems:

Perhaps the most important consequence of the pharmaceutical revolution in mental health has been to give practitioners, and therapists, a means for managing chronic mental illnesses. In this way, the pharmacological treatment of major mental illness is analogous to the use of insulin to treat diabetes. Patients with schizophrenia will still have schizophrenia, and patients with depression will still have depression, but with medication they can better deal with their illnesses, and lead more productive lives. That's a nontrivial benefit, but genuine cures for mental illness are going to have to wait for further pharmaceutical advances -- or, perhaps, another approach entirely.

Psychotherapy

Pharmacotherapy is a relatively new approach to the treatment of mental illness. Historically, the active treatment of mental illness was limited to various forms of psychotherapy, a "talking cure" in which a therapist engaged in activities that were intended to change the contents of the patient's mind: to alter patients' beliefs, feelings, desires -- and thus their behavior. Although there were important precursors, the birth of psychotherapy is commonly given as 1893, when Sigmund Freud and Joseph Breuer began publishing the articles that were eventually collected as their Studies in Hysteria.

There are literally hundreds of different psychotherapies, but they can all be classified under three major headings:

Psychodynamic Psychotherapy

Contemporary psychodynamic psychotherapy has its origins in classical Freudian psychoanalysis. But just as "neo-Freudian" theories of personality were de-biologized and de-sexualized, so has contemporary psychodynamic psychotherapy.

Jonathan Shedler (2010) has summarized the main principles of contemporary psychodynamic psychotherapy as follows:

Shedler points out that modern psychodynamic therapy, while rooted in Freud, has shed most of the trappings of classical psychoanalysis.  There is little attention paid to the Oedipus conflict and castration anxiety, for example1  Patients don't necessarily lie on a couch and free associate for 50 minutes a session, five sessions per week.  Rather, modern psychodynamic psychotherapy is practiced as an open-ended vehicle for self-examination -- which means that it might be useful even for people who don't suffer from depression, anxiety, or some other form of mental illness. 

For a first-person account by a writer who spent most of her adult lifetime (and, for that matter, a considerable amount of her childhood) in psychoanalysis, see "My Life in Therapy" by Daphne Merkin (New York Times Magazine, 08/08/2010) -- or, for a longer treatment,Mockingbird Years: A Life In and Out of Therapy (2000), a book-length memoir from Emily Fox Gordon. For some patients in long-term psychodynamic psychotherapy (and here we think of Woody Allen), the treatment -- especially if it includes hospitalization at an institution like Austen Riggs, or the MacLean Hospital, or the Menninger Clinic -- takes on a kind of Romantic feeling, not unlike the tuberculosis sanitarium depicted in Thomas Mann's novel,The Magic Mountain.

Cognitive-Behavioral Therapy

Especially where maladaptive social learning lies at the heart of the patient's illness, psychotherapy can achieve a full-fledged cure through what are essentially re-education techniques -- that is, by arranging new learning experiences that undo, or modify, the effects of old ones. Even when the patient's symptoms are not acquired through learning, at least in the usual sense, cognitive-behavioral therapy can help patients to acquire new modes of thought, and new behaviors, that will counteract the effects of their illness.

Exposure Therapies

Cognitive-behavioral therapy was first introduced for the treatment of anxiety disorders, especially phobias and obsessive-compulsive disorder. Among the most popular techniques of CBT are:

In both cases, the patient can experience complete alleviation of anxiety, which will never return -- the very definition of a cure -- through the extinction of fear, aversion, or other response, and the acquisition of more appropriate, adaptive behaviors. In the treatment of phobias, systematic desensitization and flooding are equally effective. Flooding is more efficient, but it is also somewhat more dangerous: if done improperly, the patient will be left worse off than when he started, so don't try this at home.

Relaxation Therapies

For the psychophysiological (psychosomatic) disorders, the best treatment is to eliminate the stressor from the patient's environment. However, this is not always possible, in which case the therapy involves modifying the patient's response, including physiological responses mediated by the autonomic nervous system, to the stressor.

Cognitive Restructuring

Other aspects of mental illness, particularly anxiety,depression, and the delusions of schizophrenia and bipolar disorder, appear to reflect maladaptive knowledge or belief acquired through learning.

In these cases, the goal of cognitive therapy is to change the patient's underlying cognitive structures, or schemata:

Beck's cognitive therapy for depression entails altering the patient's depressogenic schemata -- the "depressogenic triad" of negative beliefs concerning oneself, the world, and the future. Cognitive therapy also seeks to alter the patient's tendency toward arbitrary inference, selective abstraction, overgeneralization, magnification, and minimization that maintain these schemata once established.

Based on the learned helplessness theory of depression, Seligman, Abramson, Alloy, and their colleagues have suggested that depression may be alleviated by changing the individual's "depressogenic" attributional style that leads to feelings of helplessness and hopelessness, so that the patient will make more realistic, adaptive causal attributions about events.

CBT for Insomnia A good example of how behavioral and cognitive therapy can be combined is Cognitive-Behavioral Therapy for Insomnia (CBT-I).  I go into detail on this treatment because sleep is such an issue for college students -- and for so many other people, too. CBT-I has five major components: Stimulus Control: Strengthening the association between the bed and sleeping. Go to bed only when you are actually tired. Do nothing in bed except sleep and have sex. Get out of bed at the same time every morning. If you do not fall asleep within 10 minutes of going to bed, get up, move to another room, and do something relaxing until you feel sleepy. Sleep Hygiene: Altering the environment to make it more conducive to sleep. Within 4 to 6 hours of going to bed, limit the intake of substances such as caffeine, nicotine, and alcohol that interfere with sleep. Take a light snack, such as milk or peanut butter, instead. Avoid stimulating activity prior to sleep Get rid of all distractions: No TV, no computer games; Turn off the cell phone! Read a book, write an e-mail, take a warm bath. Sleep Restriction: Control the time you spend in bed, to maximize "sleep efficiency" and restore "sleep homeostasis" -- that is, the biological need to sleep. Sleep Efficiency (SE) = Total Sleep Time (TST) / Time in Bed (TIB), with the ratio expressed as a percentage. Increase TIB if SE > 90% Decrease TIB if SE < 80% Sleep Restriction involves paradoxical intention, a concept first articulated by Viktor Frankl, founder of existential psychiatry, in 1959.  It also played a prominent role in the techniques of an American maverick psychotherapist, Milton Erickson, as detailed by an early disciple, Jay Haley (Strategies of Psychotherapy, 1963; Uncommon therapy, 1973). Relaxation Training is a set of techniques, imported from systematic desensitization and stress-reduction therapies, intended to promote physical relaxation. Cognitive Therapy includes educating the patient about sleep, changing any dysfunctional attitudes or beliefs that the patient may have about sleep, and controlling the patient's worries about losing sleep.

Social Skills Training

Many behavioral disorders and annoying problems in living often result from the individual's inadequate social skills.
Public-speaking anxiety, while not necessarily rising to the level of a full-fledged social phobia, can be extremely debilitating, but can be relieved simply by giving the person practice in public speaking in a controlled, friendly environment. Many people have a great deal of trouble saying "no" to people, including their parents (or children) or spouses. In this case,assertiveness training can help people make, and respond to, demands more adaptively. People's sexual problems are not limited to impotence or frigidity. Often, the difficulties that people have in maintaining sexual arousal, or achieving orgasm (in oneself or one's partner) reflects a simple lack of knowledge about how to make love (sex is a biological function, and in that sense natural, but the pleasure-giving and -receiving aspects of lovemaking don't come naturally -- as almost anyone who remembers his or her first sexual experiences can attest. Many people learn to make love through practice, but many people need instruction in the form of sex therapy -- not necessarily from a sexual surrogate, but sometimes just some friendly advice.

Cognitive-Behavioral Therapy and Social Intelligence

Whereas pharmacotherapy attempts to alter the disordered mind by altering the chemistry of the brain, psychotherapy attempts to alter the disordered mind directly through learning experiences.

Much of cognitive-behavioral therapy seeks to alter the patient's declarative knowledge about what to believe and expect in various situations, social skills training, like relaxation training and biofeedback, seeks to alter the patient's procedural knowledge about what to do in those situations. Taken together, the cognitive-behavioral therapies work by altering the individual's social intelligence -- his or her fund of knowledge about self and others, and repertoire of interpersonal skills -- that he or she uses to navigate in the social world (see the lecture supplement on Personality and Social Interaction).

Outcomes of Psychotherapy

The notion of psychotherapy is fine in principle, and it's made a healthy living for several generations of psychiatrists, clinical psychologists, clinical social workers, and other mental health professionals. But does it really work? Psychotherapy is plagued by what might be called the Woody Allen Bugaboo -- after the characters played by the actor-director, who go through years, decades, of psychoanalysis but never seem to change.

In fact, psychotherapy came to a crisis in the 1950s when Hans Eysenck, a British psychologist, reviewed the literature and claimed that psychotherapy was ineffective -- that people who received psychotherapy had no better outcomes than those who received no treatment at all. Along with the evident success of pharmacotherapy, and lingering doubts such as expressed by the Woody Allen Bugaboo, therapists were challenged to demonstrate, scientifically, that psychotherapy really helps people with mental illness.

Since Eysenck's original study, a large body of empirical research shows that, contrary to his conclusions, psychotherapy can be an effective treatment for mental illness. For example, a classic study by Smith, Glass, and Miller (1980) employed a technique called meta-analysis (see the lecture supplement on Statistics and Methods) to combine the results of a large number of studies that compared adult patients who received various forms of psychotherapy to control patients who were untreated. Quantifying psychotherapy outcome isn't easy, but it can be done, and it's necessary if we're going to analyze the effects of psychotherapy statistically. Typically, the control patients were not denied treatment, but were merely put on a "waiting list" until a therapist became available. So, the question can be reformulated along these lines: given X months or years since their diagnosis, have patients who received psychotherapy improved more than those who did not? The answer is yes: in the Smith et al. analysis, that the median patient receiving psychotherapy did better than about 75% of the control patients.

Another finding of the Smith et al. study was that patients who received psychotherapy did better than those who did not, regardless of what form of therapy they received. This finding led some commentators to conclude that all forms of therapy are equivalent. It doesn't matter what the therapist does, so long as the patient sees one.  Saul Rosenzweig (1937) an early leader clinical psychology, has expressed this point of view as the Dodo Bird Verdict, after an episode in Lewis Carroll's Alice's Adventures in Wonderland (1865). Lester Luborsky (1975), a psychoanalyst and prominent psychotherapy researcher at the University of Pennsylvania, popularized the term.  

When Alice discovered that she would not get out of the rabbit hole, she was engulfed in a pond of her own tears, which also drenched a number of animals, such as The Mouse. The Dodo Bird suggested that the animals run a Caucus Race to get dry. The animals ran around until they ran out of breath and stopped. When Alice asked who had won the Caucus Race, the Dodo Bird replied that "everyone has won and all must have prizes".

The Dodo Bird verdict has been a source of comfort to some psychotherapists who prefer to use those psychodynamic forms of therapy that have come under attack by modern scientific psychology, as well as those who believe that the therapeutic relationship between therapist and patient is more important than anything the therapist does (e.g., Frank, 1961; Wampold, 1997, 2001)). Put another way, the "tie-score effect" is taken as indicating that the various forms of therapy have more in common than appears on the surface (see, for example, the Great Psychotherapy Debate: Models, Methods, and Findings by Bruce E. Wampold, 2001). But the Dodo Bird Verdict is troubling, too, because is suggests that the effects of psychotherapy are nonspecific -- that is, that there is no particular "active ingredient" that makes therapy effective. Put another way, it suggests that psychotherapy is simply a kind of placebo. And since psychiatric medications are approved for use precisely because they have been shown to be better than placebo, that suggests that psychotherapy is inferior to pharmacotherapy.

Fortunately, the Smith et al. study contained an analysis that showed that the Dodo Bird Verdict is not quite right: some psychotherapies work better than others. This is already evident in the data presented earlier, which showed that patients who received cognitive-behavioral forms of therapy did even better, compared to controls, than patients who received psychodynamic or humanistic forms of therapy. To examine this issue further, Smith et al. computed measures of effect size (see the lecture supplement on Methods and Statistics) for each of the studies included in their analysis.

The effect size d is a measure of the difference in mean outcomes between two treatments, expressed in standard deviation units. Thus, an effect size of 1.0 means that the average subject in the experimental group scored 1 standard deviation higher than the average subject in the control group. According to Cohen (1977), effect sizes in behavioral and social research can be classified as follows:

• d = .20: a small effect;
• d = .50: a moderate effect;
• d = .80: a large effect.

When Smith et al. calculated the average effect size for different types of therapies, all forms of therapy were shown to have at least moderate-sized effects, consistent with the Dodo Bird Verdict. However, the effect sizes associated with the cognitive and behavioral therapies were much larger than those associated with the psychodynamic and humanistic forms.

In the Smith et al. meta-analysis, the effect of psychotherapy overall, without regard to the form of therapy or the condition being treated, was quantified as an "effect size" of 0.85 -- which is generally considered a "large" effect in medical, psychological, and social-science research.

A later review by Lipsey and Wilson (1993) -- actually, a mega-analysis, or meta-analysis of published meta-analyses (18 in all), obtained a median effect size of .75 -- which is a substantial effect by any standard.

Similar results were obtained in a meta-analysis of psychotherapy for children and adolescents by Weiss and Weisz (1995).

Other studies have also demonstrated the general superiority of cognitive-behavioral therapies:

Based on these studies, and others like them, it appears that all psychotherapies are not created equal: as a rule, cognitive and behavioral therapies are more effective than psychodynamic and humanistic therapies.

Other considerations also suggest that the Dodo Bird Verdict is wrong:

For that reason, cognitive-behavioral treatments are quickly emerging as the "standard of care" in the psychological treatment of mental illness and problems in living.

This does not mean, however, that psychodynamic psychotherapy is not effective. Although classical psychoanalysis, which is what Eysenck studied in the 1950s, does not seem to be more effective than no treatment, better results have been obtained with contemporary forms of psychodynamic therapy.

For example, Shedler (2010) cited a 2006 meta-analysis of psychodynamic therapy that yielded an overall effect size of 0.97.

The real question, though, is not the comparison of overall effect sizes. Questions about the efficacy of psychotherapy are better framed more specifically:

As a rule, psychodynamic therapy appears to be based on the theory that mental illness is rooted in unconscious conflicts -- not necessarily conflicts over sex and aggression like the Oedipus conflict, but unconscious conflicts nonetheless; and that uncovering these conflicts is the key to successful therapy. However, there is very little scientific evidence that unconscious conflict lies at the root of major forms of mental illness, such as schizophrenia, anxiety disorder, or depressive disorder. And there is no scientific evidence that the primary goal of psychodynamic psychotherapy, which is to bring such conflicts into the daylight of consciousness, is the key to successful treatment. Rather, the evidence seems to indicate that the success of psychodynamic psychotherapy, where it achieves such successes, is produced by the same sorts of techniques employed by cognitive-behavioral psychotherapists -- namely, a focus on the here and now, as opposed to the there and then.

There is also the matter of utility. If psychodynamic psychotherapy achieves its positive outcomes by the same mechanisms as cognitive-behavioral therapy, but takes longer and costs more, then cognitive-behavioral therapy is to be preferred on grounds of cost-effectiveness. CBT is not just based on more scientifically valid conceptualizations of mental illness -- it also delivers more bang for the buck.

Here's an example of the kind of comparison I have in mind: a study of treatment for eating disorder (specifically, bulimia, with its characteristic cycle of binge eating and purging),  which compared psychoanalytic psychotherapy and cognitive-behavioral therapy (Poulsen et al., Am. J. Psychiat, 2014).  Patients were randomly assigned to one treatment or the other.  The psychoanalytic treatment encouraged patients to talk about threatening, repressed, feelings and desires, and how they might be related to eating disorder.  The CBT treatment challenged the patients' beliefs that their self-esteem was determined by their body weight and size, and promoted healthier eating patterns.  The psychoanalytic treatment continued for up to two years of weekly sessions, while the weekly sessions of CBT lasted only 5 months.  Nevertheless, CBT delivered clearly superior outcomes:  At the end of the 5 months, 42% of patients ceased binging and purging; treatment; the comparable figure for the psychoanalytic treatment, after the same amount of time, was only 6%.  After the full two-year treatment, only 15% of the patients in the psychoanalytic group showed remission.  There was some relapse in the CBT group, after 19 more months, but even so 44% of patients remained in remission -- a numerical increase over the outcome at 5 months. There was no untreated control group -- which is unfortunate, as spontaneous remission has been known to occur even in the case of eating disorder (Vandereycken, Eating Disorders, 2012), but even so there was a clear advantage for CBT over psychoanalysis.  Not only was CBT more effective, it was also more efficient: it took less time, and it was delivered by therapists who had less formal training than the psychoanalysts.  The study is remarkable in that the senior authors, Stig Poulsen and Susanne Lunn, were themselves psychoanalysts, and had devised the psychoanalytic treatment regime they tested.  Their treatment came out on the short end of the comparison, but they published it anyway. 

On those grounds, it seems that CBT still has an edge over psychodynamic psychotherapy. But again, a precise answer to the question will depend on the particular disorder being treated.

Psychotherapy vs. Medication

How does psychotherapy stand up, in comparison to medication?  This question has often been investigated in the context of depression.  In a provocative article entitled "Listening to Prozac But Hearing Placebo", Kirsch & Sapirstein (Prevention & Treatment,1998) conducted a meta-analysis of 19 published studies in which an antidepressant medication (such as Prozac) was compared to placebo, and another 19 studies comparing psychotherapy with a wait-list or no-treatment control.  Putting the four conditions together, they provided a comprehensive overview of the outcomes of various treatments for depression.

• The median effect size (D) for the drug groups was 1.55 standard deviations (SDs), while D for the placebo was 1.16 SDs. 
• The three types of antidepressant medication used in these studies did not differ significantly in effect size: tricyclic and tetracyclic, D = 1.52; SSRIs, D = 1.68; Other, D = 1.43. 
  
• The mean D for psychotherapy was 1.60, about the same as for drug treatment (this is a common finding), compared to D = 0.37 for the controls. 
  
• Unfortunately, K&S did not compare different forms of psychotherapy.
  

So, some depressed patients get better on their own, without any treatment -- a phenomenon called spontaneous remission.  Apparently, over the natural history of an acute episode, depression sometimes goes away all by itself.  Active treatment improves the prospects of a good outcome considerably, but the outcome with psychotherapy alone is about the same as the outcome with drugs alone.  And so is the outcome with placebo medication -- though, frankly, if you think about it, the placebo group is really another form of psychotherapy.  The patients who received placebo received all the attention, social support, and the like that the drug patients received -- only without the active drug (it's findings like this that give rise to the Dodo Bird Verdict).  On the basis of their results, K&S concluded that the active ingredients in antidepressant medications accounted for about 25% of the outcome in the drug treatment of depression; spontaneous remission (natural history) accounted for another 24%, and placebo accounted for approximately 51%.

The point of all this is not to diminish the value of antidepressants: there's that 24%, which is not nothing.   Kirsch's data is convincing, but Kramer's response is probably correct: when people are depressed, the combination of drugs and psychotherapy can be very helpful.  The point is that even powerful psychoactive drugs have substantial placebo components, and this is likely to be true for the major and minor psychedelics as well.  As Kirsch et al. (2002) point out, "Placebo alcohol produces effects that are not observed when alcohol is administered surreptitiously...". 

If pharmacotherapy is effective, and psychotherapy is effective, what about the combination of the two?Here the data is somewhat in flux, but it appears that the combination of drugs and psychotherapy is rather promising. In a study by Keller et al. (2000), depressed patients who received cognitive-behavioral therapy did about as well as those who received Serzone (Nefazodone), an SSRI, but those who received them both did especially well.
Here's another example.  Obsessive-compulsive disorder, like many other anxiety disorders, is commonly treated with SSRIs -- even though there are perfectly good cognitive-behavioral therapies for this problem.  But SSRIs aren't always effective, or at least they aren't always as effective as we'd like them to be.  In a randomized clinical trial, Simpson et al. (JAMA 2013) studied a group of patients with OCD who did not respond positively to a course of treatment with SSRIs.  One subgroup got an additional medication, the antipsychotic Risperidone (remember, "antipsychotic" drugs are really little more than major tranquilizers); another got a placebo pill; and a third group got a course of cognitive-behavioral therapy emphasizing exposure and response prevention (i.e., a variant on flooding).  The outcome was measured with the Yale-Brown Obsessive-Compulsive Scale, a standardized instrument used in the diagnosis of OCD.  The risperidone didn't help much: only about 23% of the patients got better 13% were essentially "cured"), compared to 15% (5% "cured") for placebo.  But the cognitive-behavioral therapy helped a lot: 80% of the patients in this treatment group got significantly better, and 43% were essentially "cured".  One wonders what would have happened if these patients had simply gotten the CBT, without any drugs at all.  Or, whether patients who responded well to the SSRI would have done just as well, or even better, if they received CBT as well.  Or, if those patients had just gotten CBT, without any drugs at all.

The fact of the matter is that, for people with mild or moderate cases of depression, anxiety, and many other mental illnesses, psychotherapy -- especially some form of cognitive-behavior therapy -- works about as well as medication.  Medication really boosts treatment outcome only in severely ill patients.  For people with mild to moderate illnesses, psychotherapy alone reduces side effects (obviously, because patients don't get medication in the first place) as well as the risk of relapse as the treatment proceeds and the patient is improving and recurrence of a new acute episode after full remission.

There is also evidence that psychotherapy is associated with lower relapse rates than medication. In a study of depression, patients were administered either an antidepressant SSRI or cognitive-behavioral therapy. Patients in both groups responded equally well to treatment. But when the patients were followed up some time after their medication was discontinued, or therapy terminated, the patients in the medication group were much more likely to have relapsed -- that is, to have experienced another episode of depression. In other words, psychotherapy came closer than drugs to providing a cure for depression.

And a meta-analysis by Swift et al. (Psychotherapy, 2017) found that patents who receive psychotherapy are less likely to drop out of treatment, or to refuse treatment in the first place.  These investigators analyzed studies that compared psychotherapy and pharmacotherapy, alone and in combination.  It is standard practice in these sorts of studies to report the number of patients who refused their assigned treatment, as well as the number who dropped out before completing it.  Patients assigned to pharmacotherapy alone were more likely to refuse treatment than those assigned to psychotherapy alone, or to a combination.  Similarly, patients were more likely to prematurely terminate treatment if assigned to pharmacotherapy alone.  People seeking help for mental and behavioral problems appear to prefer the personal contact that comes with psychotherapy; if they're going to get medication, they want the dialog and social support that comes with a live therapist, too.  So, even if mediation is the primary vehicle for treatment, supplying adjunctive psychotherapy may help patients stay on course.

Combining psychotherapy with medication often seems like the optimal approach to treatment.  Presently available psychiatric drugs offer a fair measure of symptom relief, but not a cure. Psychotherapy gives the patient the knowledge and skills to overcome his illness, or at least to cope with it more effectively. In the Keller study, patients who got Serzone experienced a temporary boost in their mood, and that's not a trivial outcome. But patients who got active psychotherapy learned to deal with their depression on an ongoing basis, and to adjust to the life after the depression went away. More generally, patients can suffer relapses when their drugs are withdrawn, but in a sense new knowledge and skills, acquired through the experience of cognitive-behavioral therapy, never go away. They remain permanently available to the patient, as part of his repertoire of social intelligence.

It's probably for this reason that patients who receive psychotherapy, regardless of whether or not they also receive drugs, are less likely to relapse before full remission, or to experience a recurrence of another acute episode.  On the other hand, there is some evidence that medication can actually interfere with psychotherapy (Forand et al., 2013).  For mild to moderate cases, psychotherapy does about a well as drugs -- with fewer side-effects and less risk or relapse or remission.  Drugs are most effective for the most severe cases -- and even then, psychotherapy can help, by giving the patient new knowledge and skills. 

And of course, effective psychotherapy avoids the unpleasant and harmful side-effects of medications.  A good example is insomnia, a fairly common sleep disorder, and one of the prominent symptoms of depression.  Patients with insomnia are often treated with prescription medications such as Ambien and Lunesta, as well as sedative drugs like the benzodiazepines, and over-the-counter "sleep aids" such as ZzzQuil.  But all of these drugs carry a risk of dependence, if not addiction; they can make the patient feel groggy even during the daytime; they can disrupt the REM (dreaming) stage of sleep.  But an effective psychological treatment, Cognitive-Behavioral Therapy for Insomnia (CBT-I), achieves the same success rate as medications, without the side effects (Morin et al., 1994, 2006).

The combination of drugs and psychotherapy is probably no less important in schizophrenia than it is for depression.  We know that a wide range of social stressors can be implicated in the onset of schizophrenia, and it makes sense that eliminating or at least modulating these aspects of the environment would promote successful treatment.  Equally important, people with schizophrenia may have to learn how to live with their disability, and these cognitive and social skills must be learned through active rehabilitation programs.  (See "A Social Salve for Schizophrenia" by Matthew M. Kurtz, Scientific American Mind, March-April 2013).

In the words of the ancient adage (sometimes attributed to Lao Tze in the Tao Te Ching -- though I can't find it in my copy):

Give a man a fish and he eats for a day. Teach a man to fish and he eats for a lifetime.

Giving a patient drugs is like giving a man a fish: when they're gone they're gone. But the learning that comes through active cognitive-behavioral psychotherapy stay forever, as a permanent resource for the patient.

The Social Context of Psychopathology

Traditional forms of psychotherapy, including psychodynamic and cognitive-behavioral forms of therapy, tend to treat the individual patient in isolation -- there's the therapist, and there's the patient, and that's about it. This tradition follows from the medical model, in which the patient has some illness that the doctor treats with an antibiotic, or surgery, or whatever. But psychologically speaking, we've already argued that, as the poet John Donne put it "no man is an island". Psychology explains the individual's behavior in terms of his or her individual mental states, and that goes as well for abnormal behavior, as well as for the psychological deficits and maladaptive social learning that account for it. Individuals live their lives in the context of other people, and it would be foolish to assume that the social context has no influence on individual mental patients and the course of mental illness.

In fact, we've already seen how certain anxiety disorders, such as phobias and obsessive-compulsive disorder, can be acquired through social learning, as well as through direct experience. And also how expressed emotion -- how other family members view and behavior toward the patient -- can influence the prospects for recovery and relapse in patients with schizophrenia.

The role of social factors in psychopathology can be seen in the various "epidemics" of mental illness:

Group and Family Therapy

Some therapists have gone so far as to assert that it is not enough to treat the individual patient, precisely because, in some psychologically real sense, it's not just the patient who is mentally ill. And if it's not just the patient who is mentally ill, then it's a mistake to treat the individual patient as if he or she were the only person that mattered. If the real problem is with the patient's relationship with other people, then it's the relationship that has to be treated. At the very least, other people have to be enrolled somehow in the treatment process.

This is obviously the case with many problems in living, such as marital difficulties, that are often treated by psychotherapists. If a marriage is in trouble, you can't hope to fix it by working on only one partner. Both partners have to be in the therapy, together.

Many patients are treated in groups in addition to, or instead of, individual therapy sessions.  Group therapy has obvious economic advantages, and psychological advantages as well.  Patients can learn that other people have problems like theirs, and learn how others deal with them.  Individual patients can find social support and encouragement for their own efforts to get better, and models for improvement.  Some patients' problems are best observed in a group context.  And the group provides a "safe place" where patients can try out new ideas, feelings, and behaviors.

Alcoholics Anonymous is an informal setting , created and maintained by recovering alcoholics themselves, which provides many of the benefits of group therapy.

Among the earliest and most vigorous proponents of this idea was Salvador Minuchin (Minuchin et al., 1974), a psychiatrist who has specialized in the treatment of eating disorders in adolescents. Minuchin argued that mental illness should not be construed as "contained within the individual"; nor should the mental patient be viewed as a "passive recipient of noxious environmental [or biological] influences. Rather, Minuchin argued that family (and other social) interactions may be responsible for certain forms of mental illness; and that these family interactions are truly interactional in nature, in that the patient plays a role in shaping the environment to which he or she, in turn, responds.

Minuchin's open systems model of psychopathology and psychotherapy "[broadens] the focus from the sick child to the sick child within the family" and "redefines the nature of pathological disorder and the scope of therapeutic change" (Minuchin et al., 1974). The open systems model postulates that:

Minuchin et al. conclude: "Therefore, therapy must be directed toward changing the family processes that trigger and maintain the child's... symptoms and toward changing the use of these symptoms within the family." Obviously, that can't be done with the child alone; and it can't be done by working on one family member at a time. The whole family has to be enrolled in the treatment of the child, and the whole family has to change.

Minuchin et al. go on to describe a pathological family organization in terms of several family transactional characteristics:

Minuchin et al. devised a form of family therapy that was expressly designed to identify, challenge, and break down these four characteristics, and thus create a family environment in which the child is permitted, and encouraged, to get well. For example, it wasn't just the individual child who was hospitalized for treatment; the whole family had to stop what it was doing and mobilize for treatment. Minuchin's system was further developed by a group of therapists at the Maudsley Hospital in London (yes, that's the old "Bedlam", now much reformed and one of the world's leading centers for research on psychopathology and psychotherapy), and is now known as the "Maudsley model" for family treatment of adolescent eating disorders.

In their initial 1974 study, Minuchin et al. reported about 86% success in treating 48 children with "superlabile" diabetes, "intractable" asthma, or anorexia nervosa, but they did not have a comparable group that received traditional individual therapy. The comparison of family vs. individual therapy has been carried out mostly by the Maudsley group, who largely confirmed Minuchin's findings.

Chronic Disease Management and Rehabilitation

What about instances of mental illness where a cure is impossible? There are lots of such disorders, including:

After the acute phase of mental illness, after efforts at treatment have gone as far as they can, the patient may move into a chronic phase. Such circumstances call for rehabilitation programs to help patients and others cope with their chronic disability, get out of the institution, back to their families and communities, and make an optimal social adjustment despite their illness.

Mental Hospitals

Before the 19th century, there was little by way of active treatment or rehabilitation. Psychology wasn't yet a science, nor was psychiatry a branch of medicine -- and never mind that medicine wasn't all that scientific, either!

For the most part, mental patients, when they became too much trouble, were simply warehoused -- often, kept in prisons along with criminals. Sometimes, they were housed in special "insane asylums", separate from convicts. A good example of an 18th-century insane asylum is the Royal Bethlehem Hospital in London, founded in 1337 as a religious charity, then taken under royal auspices in the 16th century. But even these hospitals could offer little more than custodial care, and conditions in most of them progressively deteriorated. Which is how the Royal Bethlehem Hospital got the nickname "Bedlam". In fact, the middle and upper classes used to pay a fee to visit the hospital and watch the antics of the patients as a form of Sunday-afternoon entertainment.

Even though Esquirol distinguished between the mentally ill (in his terms, the insane), the intellectually disabled (in his terms, the mentally deficient), and mere criminals, they were still all housed together (except for insane members of upper-class families, who were more likely to be consigned to the attic, as in Charlotte Bronte's Jane Eyre). Things began to change when Philippe Pinel (1745-1826), became the superintendent of the Bicetre, an asylum in Paris, and later the Salpetriere, a large mental hospital (which you can still see when you visit the city). Along with his mentor, Jean-Baptiste Pussin (1745-1811), Pinel pioneered the "moral" treatment of those who, although mentally ill, still deserved respect as "citizens". Pinel is also credited with freeing the mentally ill from their chains -- although it was actually Pussin who did this (and he replaced the chains with straitjackets!).

The moral, humane treatment of the mentally ill quickly spread to England and America. Bethlem Hospital was reformed. In 1792, Benjamin Rush (1745-1813), a physician who had been one of the signers of the Declaration of Independence, founded a division of the Pennsylvania Hospital (itself the first hospital in America, after New York's Bellevue), devoted to the moral treatment of the insane.  Rush's treatise, Medical Inquires and Observations Upon diseases of the Mind (1812) was the first textbook of psychiatry published in America.  Another book, On the Construction, Organization, and General Arrangements of Hospitals for the Insane (1854), by Thomas Kirkbride, the first superintendent and physician-in-chief of the Institute, remained influential even into the 20th century. (I worked at The Institute as a graduate student at the University of Pennsylvania). 

For more on Benjamin Rush, see "Rush's Remedies" by Susan Frith, Pennsylvania Gazette, 07-08/2012, and "A New Founding Mother" a sketch of Rush's wife, Julia, by Stephen Fried, Smithsonian, 2018..  Also Fried's biography of Rush, Rush: Revolution, Madness, and the Visionary Doctor Who Became a Founding Father (2018).

There was also an extensive system of private mental hospitals, catering mostly to the wealthy. An early example was the Sidis Institute at Maplewood Farms, a large estate in Portsmouth, New Hampshire. Established by Boris Sidis, a friend of William James, and leader of the "Boston School" of psychotherapy (and father to William James Sidis, at the time the youngest person ever to enter Harvard College, at age 11), the Institute offered all the latest treatments, including psychotherapy in an environment of "beautiful grounds, private parks, rare trees, greenhouses, sun parlors, palatial rooms, luxuriously furnished private baths, private farm products". By 1916, there were more than 20 such institutions in Massachusetts alone.Link to the Sidis Institute website.

Still, some of these asylums were awful places, little better than bedlam.  In 1845, the British government set up the independent Lunacy commission to set and enforce standards for private mental hospitals.  For a history of mental-hospital reform in Victorian England, see Inconvenient People by Sarah Wise (2013).

By the late 19th century publicly supported mental hospitals were a feature of virtually every state health system. These were often glorious structures, architecturally distinctive  -- following the precepts laid down by Thomas Kirkbride (see Asylum: Inside the Closed World of State Mental Hospitals by Christopher Payne (2010).

Here are some more classic state mental hospitals.

But as good and humane as these hospitals were intended to be, they still offered little more than custodial care until the beginning of the 20th century, when advances in psychology and psychiatry began to afford the possibility of active treatment of the mentally ill. In the 20th century, reflecting our progressively increased understanding of mental illness, public and private mental hospitals offered active treatments as well as custodial care: biological treatments like psychosurgery, ECT, and later drug treatments of various kinds, as well as psychotherapy by psychologists and social services by social workers. 

A 1946 expose by Life magazine described many American asylums as "little more than concentration camps".

All that came to a screeching halt with the de-institutionalization movement that began in the 1960s, when the mental hospitals began to be emptied and their residents discharged back to their families and communities. Partly this was a result of the early successes of the pharmaceutical revolution, which made many schizophrenics more manageable, and afforded symptom relief to many with patients suffering from depression and anxiety disorder. In addition, most public mental hospitals suffered from overcrowding, and budget difficulties led to a lack of properly trained and supervised staff, and corresponding scandals of the sort exposed by Life magazine.  But there were also other contributing factors:

The federal government encouraged de-institutionalization as well, for both economic and legal reasons. But the funds promised to support community and family treatment of de-institutionalized mental patients never were forthcoming -- with the result that homeless people with mental illness and substance abuse are found on the streets of every major city (if you don't believe me, check out Peoples' Park in Berkeley, or Civic Plaza in San Francisco).

For excellent journalistic coverage of the de-institutionalization movement, see:

• No One Cares About Crazy People: My Family and the Heartbreak of Mental Illness in America by Ron Powers (2018), both of whose sons suffered from schizophrenia -- and who, he argues, would have received better treatment in a mental hospital than in the community.
• Insane: America's Criminal Treatment of Mental Illness by Alisa Roth (2018), who argues that jails and prisons have become dumping grounds for the mentally ill, who receive little or no treatment for their illnesses while they are incarcerated.
  

De-institutionalization was not a bad idea. There were, admittedly, lots of people confined to mental hospitals who didn't need to be there. At the same time, it's not at all clear that the homeless mentally ill are better off on the streets of Berkeley and San Francisco than they would be within the walls of Binghamton or Napa State Hospital. Moreover, mental hospitals could have created an environment for the active rehabilitation of the mentally ill, preparing them for lives in the community and with their families.  Such environments are going to look increasingly attractive as an aging population of Baby Boomers increases the number of individuals with Alzheimer's Disease and other forms of dementia, and individuals on the autism spectrum outlive their parents and other family members who have cared for them.

And community mental health treatment isn't always all it's cracked up to be, either.  All too often, substandard care in large state institutions has been replaced by substandard care, barely better than custodial care, in small group homes, run on tight budgets, by individuals who, however well-intentioned they may be, often lack proper training and supervision.  And then again, many of these group homes are run on a for-profit basis, further limiting the care and support that the non-institutionalized mentally ill receive.  Many homeless mentally ill and up in local jails, for want of any other way of housing them -- a step backward, I suppose, toward the moral model of mental illness.

For this reason, some prominent psychiatrists have begun to advocate for a revival of the state mental hospital system for the care of the chronically mentally ill (see "Improving Long-Term Psychiatric Care: Bring Back the Asylum" by D.A. Sisti, A.G. Segal, and E.J. Emanuel, Journal of the American Medical Association, 2015).  The same legitimate concerns that led to de-institutionalization in the first place will probably oppose any such move (see, for example, "Under Lock & Key: How Long?" by Aryeh Neier & David J. Rothman, New York Review of Books, 12/17/2015).  But if re-institutionalization occurs, we can hope that it avoids the problems that cropped up in the 20th century.  That will take money, for proper facilities and proper staffing; and strict enforcement of the legal principle of "least restrictive treatment".  The mental hospital may have its place in public policy, but nobody should be sent there who can't receive safe, effective treatment in the community.

For better and for worse, de-institutionalization is a fact of life in the United States, and many other developed countries as well.  But in underdeveloped countries, institutionalization remains the norm, and the institutions themselves are often little more than warehouses for the mentally ill.  In 2013, the World health Organization introduced a "global mental health plan" to move from centralized mental hospitals to community-based care.  But this takes money -- money that is hard to come by even in a rich country like the United States, and even harder to come by for the underdeveloped countries of the Third World.  (For more information, with a special focus on developments in Guatemala, see "Where Mental Asylums Live On" by John Rudolf, New York Times, 11/05/2013.)

Token Economies

Even in the case of the organic brain syndromes, developmental disorders, and functional psychoses, behavioral treatments such as token economies, employing the principles of instrumental or operant conditioning, can facilitate the acquisition of new, more adaptive behaviors in individuals with actual or presumed brain pathology. In token economies, patients receive tokens, such as poker chips, contingent on their performance of certain actions, such as cleaning their living area or dressing themselves; these tokens may then be exchanged for goods at the hospital canteen or other privileges. In the language of instrumental conditioning, then, the tokens are secondary reinforcers. Perhaps as much as medication, these training procedures help chronic mental patients return to their families, and live in the community, by providing them with the repertoire of behaviors they need to live outside the confines of the mental hospital. (Token economies can also serve as laboratory models of national economies, but for some reason behavioral economists haven't taken much interest in them.)

Social Interventions for Autism

Autism is one of the most challenging chronic mental illnesses.  Psychotropic medication can help manage some of the secondary symptoms of autism, such as the mood swings, temper tantrums, and irritability that autistic children display, but have essentially no impact on the child's primary deficits in communication and other aspects of social interaction.  For that, the best hope lies in psychosocial interventions (though injections of oxytocin have been recommended!).

One intervention technique, derived from instrumental conditioning, is applied behavior analysis (ABA), first devised by O. Ivar Lovaas for the treatment of children with autism. This treatment program involves an intensive regime, perhaps 40 hours a week (i.e., a full-time job for both the child and his therapists), beginning very early (as young as 3 years of age). ABA involves an extensive system of rewards and punishments intended to shape and reinforce desirable behaviors, such as making eye contact and sitting quietly, and to eliminate or discourage undesirable ones, like yelling or head-banging. Early applications of ABA sometimes used punishments such as slapping or even electrical shocks (delivered by an instrument that looked disconcertingly like a cattle prod), resulting in considerable criticism. Yet Lovaas contended that his methods delivered results when other strategies did not, and he eventually gave up punishment in favor of a regime based entirely on positive reinforcement (in this respect, following Skinner's dictum that positive reinforcement is much better than punishment for controlling behavior). In 1987, Lovaas reported that his treatment had a 50% success rate (a 1993 follow-up showed that most of the patients had maintained their treatment gains). Subsequent studies have not reported about a 30% success rate, which is still better than most other treatments for autism. ABA is phenomenally expensive, unless you consider the alternatives, and it is now a standard treatment for autism.

A number of other interventions have combined techniques from ABA with more cognitive approaches derived from the study of social development and the "theory of mind".  For example, the Early Start Denver Model (ESDM) focuses on getting the child to pay attention (and respond appropriately) to social cues such as facial expressions, gestures, and -- last but not least -- speech. Studies have shown that ASDM is effective, but it isn't easy, one study employing more 2000 hours of therapy delivered over the course of two years.

Some variant of ABA is the most popular, and effective, approach to treatment currently available for autism spectrum disorders (ASD).  There is no medication available, other than sedatives.  And although we generally think of autism as a "chronic" mental illness, from which a patient will never fully recover, in fact the prospects for improvement are not trivial.  As with schizophrenia, it is possible for an autistic individual to make a substantial-enough recovery that he or she will no longer qualify for ASD. 

One conclusion from these studies is that, as with schizophrenia, prognosis is related to premorbid personality as well as to active treatment.  Those patients who do better, achieving an optimal outcome of their illness, may be those who have better resources at the start: better social skills, higher intelligence, and the like.

Cognitive Restructuring -- Again

As discussed earlier, a number of theorists have proposed that paranoid delusions reflect schizophrenic patients' inappropriate attempts to explain the anomalous experiences they have as a result of their illness. Accordingly, another aspect of rehabilitation may be to give delusional patients more appropriate and adaptive explanations for what is happening to them.

When I visited China in 1985, as part of a delegation of mental health specialists, the mental-health authorities we meat stressed that the incidence of schizophrenia in the People's Republic of China was no different than that in other countries -- which is what we would expect if schizophrenia was due largely to genetic and biochemical factors. However, we were also informed that the incidence of paranoid schizophrenia was lower in China than in other countries. I have no independent knowledge of whether this is so, but the attributional account of delusions offers an explanation for this fact (assuming it is true): in China, social organization is such that the mentally ill are detected very early in the acute phase of their illness, and are brought to local mental hospitals for treatment. At least at the time of my visit, it was routine for these acute patients to be given antipsychotic medication followed by a series of lectures on the nature of their illness. Perhaps, when schizophrenics in the acute stage of their illness are given correct information about what is happening to them, they have no need or opportunity to develop delusional explanations for themselves.

Rehabilitation is an important aspect of mental-health treatment because, at least for the present, most serious mental illnesses, such as schizophrenia, autism, and the more serious forms of mood disorder, are chronic diseases: psychiatric medications provide only symptomatic relief, and psychotherapy can only go so far. For mental health, as for the rest of medicine, in the face of incurable illness we do not simply throw up our hands and walk away from the patient. Nor, for that matter, do we continue fruitless attempts to cure the patient's illness. Rather, in chronic illness the goal of treatment shifts from cure to rehabilitation -- to help the mentally ill become at least partly self-sufficient, to live on their own or with their families, or in halfway houses and other protected living environments. In mental health, as in the rest of medicine, permanent hospitalization is the last resort.

Stigma, Stereotypes, and the Self-Fulfilling Prophecy

Unfortunately, the prospects of successful treatment of mental illness -- whether success comes in the form of a cure or rehabilitation -- is hampered by the social context in which mental illness occurs.

More subtle, perhaps, is the frequent occurrence of stereotyping when it comes to the mentally ill, as well as the dominance of first impressions -- psychiatric diagnoses, once made, tend to stick so that the person never sheds the label of "schizophrenic", "manic-depressive", etc., or for that matter the sick or impaired role. Generally, there is a popular refusal to admit the possibility of a good outcome in mental illness -- a successful treatment, whether cure or rehabilitation, that would allow the person to leave the sick role and assume his or her proper role(s) in society.

Erving Goffman, a sociologist, analyzed what he called the stigma of mental illness. For Goffman, a stigma is "an attribute that is deeply discrediting" -- which turns a "whole person" into a "tainted, discounted one". Many physical stigmata (that's the plural of stigma) are immediately apparent. But others, like the stigma of mental illness, are not readily apparent. The mentally ill only become stigmatized when their mental illness becomes known to others. Before their conditions become known, the mentally ill, with their secret stigma, are discreditable; after their condition becomes known, the mentally ill are actually discredited.

Following Goffman, Jones and his colleagues analyzed the stigma of mental illness in terms of a number of different dimensions.

Link and Pheelan (2001) offered a different perspective on the components of the stigma of mental illness.

Considerations of stigma and stereotyping lead us back to our earlier discussion of the various construals of deviance.

Also subtly keeping the mentally ill "sick" is the self-fulfilling prophecy (discussed in the lecture supplements on Personality and Social Interaction):

Such a process can lead patients to define themselves as incurably ill, diminishing their motivation for therapeutic change. It can also lead those who care for mental patients to substitute custodial care and medication for active treatment that might return patients to their normal role(s) in society. If there are few or no attempts at cure or rehabilitation, we can virtually guarantee that mental patients will never get well.

A good example comes from schizophrenia, which is generally thought to have a poor prognosis.  In fact, people with schizophrenia can show a remarkably good recovery, with treatment, so long as they get the right treatment, in the right environment  -- and, perhaps, have a "better" premorbid personality to begin with.

Yet another approach to the stigma of mental illness is simply to deny it.  Some individuals with autism, for example, deny that autism is a mental illness which should be treated an eliminated.  Instead, they argue that autism exemplifies neurodiversity.  In this view, autistic individuals are not mentally ill -- they just have brains that operate differently than most other people.  And while they may need help and accommodation in some respects, they assert that autism isn't something to be eliminated.  Rather, they incorporate autism into their personal identities. 

The "Pseudopatient" Study

The deleterious effects of the social context on the treatment of the mentally ill are illustrated by a controversial study reported by David Rosenhan, a professor of psychology and law at Stanford University, in 1973. In this study, Rosenhan, as well as some colleagues and students, sought and gained admission to a number of different mental hospitals, public and private, by falsely reporting some symptoms commonly associated with mental illness. All the "pseudopatients" were in their mid-30s (except perhaps for Rosenhan himself), gainfully employed, with no prior history of psychopathology, and during the admission interview all the patients told the truth about themselves except for two matters:

Given that hallucinations are serious symptoms, it is not surprising that all of the pseudopatients were admitted to the hospital for observation. What is surprising is what happened next.

Immediately upon their admission, the pseudopatients ceased their simulation, and behaved normally in every way (except for identifying themselves explicitly as simulators). Other patients on the ward frequently noticed the change, but by and large the professional staff did not. In fact, Rosenhan reports that the pseudopatients were largely ignored by the staff.

A possibly apocryphal story: In one particular episode, Rosenhan spent a sabbatical quarter as a pseudopatient. At the end of the term, when Rosenhan was obliged to return to his teaching duties, he informed the attending psychiatrist that he was Prof. David Rosenhan of Stanford University. The response was "Oh,sure you are!". Rosenhan's wife had to secure a legal write of habeas corpus to get him discharged.

And another one. In another episode, one of Rosenhan's colleagues, an international authority on depression, while masquerading as a pseudopatient, noticed that a particular patient was being treated for depression with an inappropriate drug. The colleague, while staying in his pseudopatient role, approached one of the ward psychiatrists to discuss the matter -- identifying himself merely as someone who had read a lot about depression. After the discussion, the psychiatrist made a notation on the pseudopatient's chart that he displayed "grandiosity", and increased his medication!

A cautionary note:  In 2019, Susannah Cahalan published a book, The Great Pretender: The Undercover Mission that Changed Our Understanding of Madness, which was a journalistic inquiry into the pseudopatient study.  Cahalan's previous book, Brain on Fire: My Month of Madness (2016), was an account of her own psychotic episode, resulting from a neurological autoimmune disorder -- a form of encephalitis initially misdiagnosed as a mental illness.  Going through Rosenhan's papers (he died in 2012), Cahalan was  largely unable to track down the other pseudopatients, or the raw data for the study he published in Science, and she strongly implied that he may have fabricated his data. 

That Rosenhan and his collaborators were admitted to mental hospitals is unremarkable. Anyone who reports auditory hallucinations deserves some further investigation. But after their admission, the treatment of the pseudopatients can only be described as gross negligence. There was little or no investigation of the "presenting complaints" that brought them to the hospital in the first place, and the clinical staff failed to notice that their symptoms had "remitted". There was little or no active attempt at cure or rehabilitation, or apparently any consideration that active treatment was possible.

Although the pseudopatient study is often cited as an example of the negative effects of the medical model in psychiatry, it would be more accurate to say that the problems encountered by the pseudopatients occurred precisely because the clinicians failed to adhere to the medical model. If the psychiatrists and others had acted in accordance with the medical model, they would have discovered much sooner that the patients' symptoms had disappeared; they would have been more observant of their behavior; and they would not have been so quick to dispense medication to patients who did not need it.

The issue of how we label people with mental illness has come to the fore with the "disability rights" movement, and the objection of people who have various disabilities to be identified with their disabilities (a similar issue has been raised in racial, ethnic, and sexual minority communities as well).

One important question is how to refer to people with various disabilities.  Put bluntly, should we say that "Jack is a schizophrenic or "Jack is a person who ha schizophrenia"?  Or substitute any other diagnostic label, including neurotic, depressive, or autistic. 

Dunn and Andrews (American Psychologist, 2015) have traced the evolution of models for conceptualizing disability -- some of which also apply to other ways of categorizing ourselves and others.  The current debate offers two main choices:

Mental-Health Policy

Until recently, the treatment of the mentally ill was left pretty much in the hands of physicians, with minimal regulation. Historically, legislatures and courts have not intruded on issues of medical treatment -- relying, implicitly on physicians' Hippocratic Oath to "do no harm" -- and also out of respect for the sapiential authority of physicians, who are assumed to have more expertise in matters of diagnosis and treatment than laypeople do.

This situation changed sharply in 1971, in Wyatt v. Stickney, a landmark class-action case brought against the Alabama state mental hospitals in 1971. Ricky Wyatt (1954-2011) had a record of youthful misbehavior, as a result of which his juvenile probation officer arranged to have him committed to Bryce State Hospital at the age of 14 -- the youngest patient there (the procedures for such institutional commitment were pretty lax at the time). Despite the fact that he never actually received a psychiatric diagnosis, Wyatt was "treated" with large doses of Thorazine and other antipsychotic medications, and suffered many other indignities. The Federal judge in the case, Frank Johnson, ruled in favor of the plaintiffs and placed the entire state hospital system under federal receivership (where it stayed until 2003). He also issued a set of guidelines for the proper treatment of mental patients, now known as the "Wyatt Standards", that are now applied nationwide. Chief among these is the concept of least restrictive treatment -- that if mentally ill or intellectually disabled persons must be institutionalized, they have a right to as much freedom as practicable. They also have a right to human treatment, sufficient staffing, and individualized treatment plans, plus certain minimal standards for diet and nutrition. It is no accident that the same judge who ruled in Wyatt had also earlier placed Alabama's schools and prisons under federal receivership. The case is a landmark of civil rights law, and the Wyatt Standards are sometimes known as the Mental Patients' Bill of Rights.

Another major change in mental-health policy occurred in 1999, with the White House Conference on Mental Health and the issuance of the Surgeon General's Report on Mental Health. The Surgeon General argued that "mental health is fundamental to health" -- that a sound mind is part and parcel of a sound body. It also stressed that "mental health disorders are real health conditions", not figments of someone's imagination or excuses for not working. It asserted that "the efficacy of mental health treatments is well documented" -- a real change from the Eysenck study of the 1950s, and the Woody Allen Bugaboo. And it noted that "a range of treatments exists for most mental disorders", including both biological treatments (like drugs) and psychotherapy. This was the first time that federal policy formally recognized the problem of mental illness.

Mental Health Parity

Before this time, mental illness was treated quite differently from physical illness. While many consumers has insurance policies like Blue Cross and Blue Shield to help pay medical bills, they had to pay for psychotherapy out of their own pockets. And even "Cadillac" health-insurance plans imposed annual or lifetime dollar limits on expenditures for treatment for mental illness and substance abuse. For example, my own health policy at the University of California, pays for only 28 days of inpatient mental-health treatment, and outpatient psychotherapy had to be authorized in a way that outpatient medical treatment did not.

Now, however, by federal law, such as the Mental Health Parity Act of 1996, there is parity between "medical" and "mental" illness.

Still, there remained important gaps between mental health and other medical services.  In particular, the MHPA control issues like cost-sharing, limits on number of therapy visits or days of inpatient hospitalization, and the like.  As a result, employers and insurance companies were able to circumvent the intention of the Act by increasing patient co-pays for mental-health services, and imposing limits on the number of visits or the days of coverage.    

The Mental Health Parity and Addiction Equity Act of 2008 was intended to strengthen parity even further, by closing the loopholes in the MHPA.  It required that all financial requirements, including co-payments and caps on visits or days of treatment be the same for mental-health services as for medical and surgical services.
So did the Affordable Care Act of 2010 (aka ObamaCare) reinforced parity -- in fact, one Administration official was quoted as saying that it was "kind of the final word on parity" ("Rules to Require Equal Coverage for Mental Ills" by Jackie Calmes and Robert Pear, New York Times, 11/08/2013). 

Again, any requirements and restrictions (such as pre-authorization for treatment) must be the same for mental health as for other medical care.  Still, talk (legislation) is cheap: the real issue now is enforcement, in which other budgetary priorities may conspire with the stigma associated with mental illness, and skepticism about the value of mental-health treatments, to prevent mental-health services from actually achieving parity. 

In fact, there are serious challenges to parity for mental-health and substance-abuse services.   In 2017, President Trump and the Republican-controlled House and Senate attempted to "repeal and replace" the Affordable Care Act (ACA, aka "Obamacare").  Every attempt to repeal and replace failed, but by the smallest of margins -- just one vote in the Senate would have meant major changes to the funding of mental-health and substance-abuse treatment. 

• One reason is that one of the "reforms" proposed by the Republicans was to allow individual states to waive the requirement that certain "essential services" be covered by health-insurance policies.  Among those "essential services" most threatened were pregnancy and maternity care, and -- you guessed it -- mental-health and substance-abuse treatment.  The rationale offered in all three cases was that because such services were not likely to be needed by everyone -- only half the population can become pregnant, for example! -- then only those in need of such services should have to pay for insurance to cover them.  The argument is bogus, of course:  the whole point of insurance, of any kind, including auto and homeowner insurance, is to spread the risk, and the costs, among people who may not need it -- those who may never have an auto accident, for example, or whose homes might not burn down.  But it's telling that (almost) the first thing the Republicans promoting health-care "reform" thought of was to eliminate was parity for mental-health and substance-abuse services.
• Another reason is that each of the Republican proposals loosened the requirement, under the ACA, that individuals with pre-existing conditions not be charged more for health insurance than those without.  Individuals already diagnosed with heart-disease or cancer could not be denied insurance, under these "reform" plans; but they could be charged more to be insured against the costs of treatment.  The problem this raises, with respect to mental-health coverage, is that many mental illnesses, such as depression, are subject to relapse and recurrence.  If you've already had an episode of depression, even if you've fully recovered, that may count as a "pre-existing condition" that will increase your insurance rates.  Moreover, remember that in the diathesis-stress model of mental illness, some individuals may demonstrate a relatively poor "pre-morbid adjustment" even before their first episode.  That, too, may be counted as a pre-existing condition, leading to increased insurance rates.
  

Similar issues arise at the state level. 

In 1999, California passed the California Mental Health Parity Act to implement and extend the federal MHPA.

Despite the movement toward parity, many psychiatrists claim that they are not paid enough for their services.  As a result, they do not accept insurance, and require that their patients pay out of pocket.  A 2013 study by Bishop et al. (JAMA Psychiatry) found that only 55% of psychiatrists accepted private insurance, compared with 89% of other physicians, and only 55% of psychiatrists accept Medicare patients, compared to 86% of other specialists, and only 43% accept Medicaid patients, compared to 73% of their colleagues.  Whether these disparities reflect a problem with reimbursement for mental-health services, or the mendacity of some mental-health professionals, isn't clear.  mental-health.  But it makes clear that mental-health parity is only part of the problem of delivering mental-health services.  In addition to balking at (allegedly) low payments, these psychiatrists may also resist the kind of intrusive review that comes with third-party payments.  To be fair, many psychiatrists are solo practitioners, and simply don't have the office staff needed to cope with the paperwork that comes with insurance -- and solo practitioners in all specialties are less likely to take insurance than are physicians in any sort of group practice.  Then again, psychotherapy takes time, and so long as insurers (including Medicare and medicaid) reimburse on a fee-for-service basis, psychotherapists will never be able to move as many patients through their practice, or deliver as many services per unit time, as other physicians (or, for that matter, psychiatrists who do little more than dispense medication).  

Perhaps for this reason, there aren't enough psychiatrists and other mental-health professionals to meet the demand.  A 2012 study by the US Department of health and Human Services found that fully 55% -- that's more than half -- of the nation's 3100 counties have no practicing psychiatrists, psychologists, or social workers. 

And mental-health workers don't earn as much as other medical professionals.  A 2012 survey by Medscape showed that the average income for psychiatrists is only $186,000/year, ranking psychiatry 19th out of 25 medical specialties.  Now, $186,000 is not chump change -- but it's not nearly as much as other physicians make.

So it's a system, and it can't be fixed simply by mandating parity.  Mental-health professionals don't get paid enough, so there aren't enough of them. 

Still, if psychotherapists could show that what they do really works, then insurers would have little choice but to pay for the services they deliver.  Which brings us to the most important feature of the current mental-health environment, which is the move toward evidence-based practices. 

Evidence-Based Practice

But these changes in mental-health policy and practice come with a price, which is that, for the first time, mental-health practitioners have to demonstrate that they know what they're doing -- that their diagnoses are valid, and that their treatments work.

Frankly, up until these changes in policy, people didn't care much what went on in psychotherapy -- largely because they weren't paying for it. People can do what with their own money, the thinking went -- they can buy cigarettes, or speedboats, or psychotherapy. But if they're going to spend my money, whether in the form of insurance premiums or taxes, they better be spending it on something they really need, like a valid diagnosis, and something that really works, like an effective treatment. And also, frankly, "third-party" payers were still suspicious that mental illness was a bogus concept, and psychotherapy a bogus treatment. If people wanted to waste their own money on such self-indulgence, that was fine. But they weren't going to waste mine.

The upshot of all of this is that mental-health parity was a huge boon for mental-health practitioners, because they became eligible for third-party payments from government and insurance companies -- not to mention that parity also helped reduce the stigma of mental illness. But parity came with a price -- which is that mental-health practitioners actually had to prove, for the first time, that their treatments actually worked.

But of course, the very question of EBPs suggests that there are some treatments that don't work, and some practices that aren't empirically valid. Could this possibly be true?The simple, straightforward answer is "Yes" -- but this is also true for medicine in general, not just psychotherapy. The medical profession has long cloaked itself with the mantle of science, but until relatively recently physicians had relatively few effective treatments for disease. Mostly, their treatments were palliative in nature, intended to ameliorate the patient's symptoms, and make the patient comfortable, while nature took its course; or else they simply removed diseased organs and tissues through surgery. Scientific medicine really only began with the microbe-hunting of Louis Pasteur and Robert Koch in the 19th century, and successive phases of the pharmaceutical revolution of the 20th century. It is only relatively recently that medical researchers have begun to test medical practices to determine whether they actually work, which ones work better than others, and which are cost effective. Evidence-based medicine is epitomized by the clinical trials that new drugs must go through, to demonstrate their safety and efficacy, before they are marketed for the treatment of specific diseases.

Something similar is now happening to psychotherapy. For a long time, psychotherapists, including psychiatrists and clinical social workers as well as clinical psychologists, have had to operate "in the dark" about whether their treatments were actually effective. Many psychotherapists were loathe to measure the effects of their treatments quantitatively. And so long as patients were paying out of their own pockets, and so long as they believed that they were being helped by their therapists, there was little incentive for psychotherapists to validate their practices scientifically. In the 20th century, as standards for medical practice changed, psychotherapists felt under increasing pressure to demonstrate that their practices, too, "really worked". The pressure increased as responsibility for paying for psychotherapy gradually shifted to "third parties" (patients and their therapists were the first an second parties) such as employers and health-insurance firms. These third-party payees naturally want to make sure that they were getting value for their money: and so they demanded that psychotherapists, like other health professionals, show that their practices were both effective and cost-effective. Although there is an increasing literature on the validity of various assessment practices, such as comparing "objective" and "projective" techniques (see, e.g., "Clinical Assessment" by J.M Wood, H.N. Garb, S.O. Lilienfeld, M.T. Nezworski,Annual Review of Psychology, 2002), most practice research focuses on treatment practices.

And, for that matter, something similar is happening in other fields of public policy.  Under the Obama Administration, the federal Office of Management and Budget has tried to promote experiments structured like clinical trials, with random assignment of subjects (or, for that matter, organizations) to conditions, as a rational basis for policy changes.  This trend is most advanced in the field of education.  Beginning with the passage of the No Child Left behind Act during the George W. Bush (43) Administration, the US Department of Education has posted the findings of "clinical trials" of educational innovations on its What Works Clearinghouse website.  Included among these innovations are many of the teaching and learning strategies discussed in the Exam Information page.

Another name for evidence-based treatments is "empirically supported treatment", or EST. The term "evidence-based practice" (EBP) extends the scope of the EST movement to procedures employed for diagnosis and assessment.

At present, the standards for evidence-based practice in psychotherapy are roughly modeled on the clinical trials required before drugs are marketed (see, e.g., "Empirically Supported Psychological Interventions: Controversies and Evidence" by D.L. Chambless & T.H. Ollendick,Annual Review of Psychology, 2001). In order to qualify as "empirically supported", a treatment must yield outcomes that are significantly better than those associated with an adequate control (typically, patients who receive no treatment at all) in at least two studies, preferably conducted by independent research groups. An ongoing list of treatments that meet current standards is maintained by Division 12 (Clinical Psychology) of the American Psychological Association.

There is a legitimate debate within and between the science and clinical communities about what constitutes proper standards for ESTs.

Right now, the standards for EST are pretty minimal. They are a good start, but the standards need to be ratcheted up (the opposite of dumbing down, I guess) over time to improve the quality of psychotherapeutic practice.

It would represent a major "ratcheting up" of the EST requirements if we required not just that the treatment be efficacious, but that the theory on which the treatment is based be scientifically valid as well. But it may be exactly what we need (see below for more on this subject).

These are all points for debate within the field. The actual standards for EST are the result of a political process, and inevitably involve some compromise. But once they are established, the important thing is that (1) they are adhered to and (2) that if they are changed the change is in the direction of tightening not loosening. If practice is to be based on science, and science goes forward, there can be no going back for practice.

It took a Supreme Court case, Virginia Academy of Clinical Psychologists, and Robert J. Resnick, Ph.D. v. Blue Shield of Virginia, (1977), to establish psychologists' legal right to practice, and receive reimbursement, without being supervised by psychiatrists or other physicians.  Clinical psychology owes is autonomy from psychiatry, and its eligibility for third-party payments, to the assumption that its practices rest on a firm scientific foundation. Therefore, clinical psychology, and the rest of the mental-health profession, departs from scientific evidence at its own risk.

Putting Mental Illness in Social Context

Rosenhan's "pseudopatient" study illustrates an important general point about mental illness, and for that matter about normal mental and behavioral functioning as well.

What is true for normal mental functioning is true for mental illness as well. We simply cannot treat mental illness successfully by operating on the individual alone. We must also act to change the social environment in which the person lives.

For therapy to succeed, mental-health professionals must do more than address the individual's underlying psychopathology -- psychological deficits, maladaptive social learning, diathesis and stress, biological substrates, and so on.