Overview of Etiology

The precise causes (etiology) of most mental disorders are not known. But the key word in this statement is precise. The precise causes of most mental disorders—or, indeed, of mental health—may not be known, but the broad forces that shape them are known: these are biological, psychological, and social/cultural factors.

What is most important to reiterate is that the causes of health and disease are generally viewed as a product of the interplay or interaction between biological, psychological, and sociocultural factors. This is true for all health and illness, including mental health and mental illness. For instance, diabetes and schizophrenia alike are viewed as the result of interactions between biological, psychological, and sociocultural influences. With these disorders, a biological predisposition is necessary but not sufficient to explain their occurrence (Barondes, 1993). For other disorders, a psychological or sociocultural cause may be necessary, but again not sufficient.

As described in the section on modern neuroscience, the brain and behavior are inextricably linked by the plasticity of the nervous system. The brain is the organ of mental function; psychological phenomena have their origin in that complex organ. Psychological and sociocultural phenomena are represented in the brain through memories and learning, which involve structural changes in the neurons and neuronal circuits. Yet neuroscience does not intend to reduce all phenomena to neurotransmission or to reinterpret them in a new language of synapses, receptors, and circuits. Psychological and sociocultural events and phenomena continue to have meaning for mental health and mental illness.

Much of the research that is presented in the remainder of this report draws on theories and investigations that predate the more modern view of integrative neuroscience. It is still meaningful, however, to speak of the interaction of biological and psychological and sociocultural factors in health and illness. That is where the overview of etiology begins—with the biopsychosocial model of disease, followed by an explanation of important terms used in the study of etiology. Then, against the backdrop of the introductory section on brain and behavior, the following sections address biological and psychosocial influences on mental health and mental illness, a separation that reflects the distinctive research perspectives of past decades. The overview of etiology draws to a close with a discussion of the convergence of biological and psychosocial approaches in the study of mental health and mental illness.

Biopsychosocial Model of Disease

The modern view that many factors interact to produce disease may be attributed to the seminal work of George L. Engel, who in 1977 put forward the Biopsychosocial Model of Disease (Engel, 1977). Engel’s model is a framework, rather than a set of detailed hypotheses, for understanding health and disease. To many scientists, the model lacks sufficient specificity to make predictions about the given cause or causes of any one disorder. Scientists want to find out what specifically is the contribution of different factors (e.g., genes, parenting, culture, stressful events) and how they operate. But the purpose of the biopsychosocial model is to take a broad view, to assert that simply looking at biological factors alone—which had been the prevailing view of disease at the time Engel was writing—is not sufficient to explain health and illness.

According to Engel’s model, biopsychosocial factors are involved in the causes, manifestation, course, and outcome of health and disease, including mental disorders. The model certainly fits with common experience. Few people with a condition such as heart disease or diabetes, for instance, would dispute the role of stress in aggravating their condition. Research bears this out and reveals many other relationships between stress and disease (Cohen & Herbert, 1996; Baum & Posluszny, 1999).

One single factor in isolation—biological, psychological, or social—may weigh heavily or hardly at all, depending on the behavioral trait or mental disorder. That is, the relative importance or role of any one factor in causation often varies. For example, a personality trait like extroversion is linked strongly to genetic factors, according to identical twin studies (Plomin et al., 1994). Similarly, schizophrenia is linked strongly to genetic factors, also according to twin studies (see Chapter 4). But this does not mean that genetic factors completely preordain or fix the nature of the disorder and that psychological and social factors are unimportant. These social factors modify expression and outcome of disorders. Likewise, some mental disorders, such as post-traumatic stress disorder (PTSD), are clearly caused by exposure to an extremely stressful event, such as rape, combat, natural disaster, or concentration camp (Yehuda, 1999). Yet not everyone develops PTSD after such exposure. On average, about 9 percent do (Breslau et al., 1998), but estimates are higher for particular types of trauma. For women who are victims of crime, one study found the prevalence of PTSD in a representative sample of women to be 26 percent (Resnick et al., 1993). The likelihood of developing PTSD is related to pretrauma vulnerability (in the form of genetic, biological, and personality factors), magnitude of the stressful event, preparedness for the event, and the quality of care after the event (Shalev, 1996).

The relative roles of biological, psychological, or social factors also may vary across individuals and across stages of the life span. In some people, for example, depression arises primarily as a result of exposure to stressful life events, whereas in others the foremost cause of depression is genetic predisposition.

Understanding Correlation, Causation, and Consequences

Any discussion of the etiology of mental health and mental illness needs to distinguish three key terms: correlation, causation, and consequences. These terms are often confused. All too frequently a biological change in the brain (a lesion) is purported to be the“cause” of a mental disorder, based on finding an association between the lesion and a mental disorder. The fact is that any simple association—or correlation—cannot and does not, by itself, mean causation. The lesion could be a correlate, a cause of, or an effect of the mental disorder.

When researchers begin to tease apart etiology, they usually start by noticing correlations. A correlation is an association or linkage of two (or more) events. A correlation simply means that the events are linked in some way. Finding a correlation between stressful life events and depression would prompt more research on causation. Does stress cause depression? Does depression cause stress? Or are they both caused by an unidentified factor? These would be the questions guiding research. But, with correlational research, several steps are needed before causation can be established.

If a correlational study shows that a stressful event is associated with an increased probability for depression and that the stress usually precedes depression’s onset, then stress is called a “risk factor” for depression.⁵ Risk factors are biological, psychological, or sociocultural variables that increase the probability for developing a disorder and antedate its onset (Garmezy, 1983; Werner & Smith, 1992; Institute of Medicine [IOM], 1994a). For each mental disorder, there are likely to be multiple risk factors, which are woven together in a complex chain of causation (IOM, 1994a). Some risk factors may carry more weight than others, and the interaction of risk factors may be additive or synergistic.

Establishing causation of mental health and mental illness is extremely difficult, as explained in Chapter 1. Studies in the form of randomized, controlled experiments provide the strongest evidence of causation. The problem is that experimental research in humans may be logistically, ethically, or financially impossible. Correlational research in humans has thus provided much of what is known about the etiology of mental disorders. Yet correlational research is not as strong as experimental research in permitting inferences about causality. The establishment of a cause and effect relationship requires multiple studies and requires judgment about the weight of all the evidence. Multiple correlational studies can be used to support causality, when, for example, evaluating the effectiveness of clinical treatments (Chambless et al., 1996). But, when studying etiology, correlational studies are, if possible, best combined with evidence of biological plausibility (IOM, 1994b).⁶ This means that correlational findings should fit with biological, chemical, and physical findings about mechanisms of action relating to cause and effect.

Biological plausibility is often established in animal models of disease. That is why researchers seek animal models in which to study causation. In mental health research, there are some animal models—such as for anxiety and hyperactivity—but a major problem is the difficulty of finding animal models that simulate what is often uniquely human functioning. The search for animal models, however, is imperative.

Consequences are defined as the later outcomes of a disorder. For example, the most serious consequence of depression in older people is increased mortality from either suicide or medical illness (Frasure-Smith et al., 1993, 1995; Conwell, 1996; Penninx et al., 1998). The basis for this relationship is not fully known. The relationship between depression and suicide in adolescents is presented in Chapter 3.

Putting this all together, the biopsychosocial model holds that biological, psychological, or social factors may be causes, correlates, and/or consequences in relation to mental health and mental illness. A stressful life event, such as receiving the news of a diagnosis of cancer, offers a graphic example of a psychological event that causes immediate biological changes and later has psychological, biological, and social consequences. When a patient receives news of the cancer diagnosis, the brain’s sensory cortex simultaneously registers the information (a correlate) and sets in motion biological changes that cause the heart to pound faster. The patient may experience an almost immediate fear of death that may later escalate to anxiety or depression. This certainly has been established for breast cancer patients (Farragher, 1998). Anxiety and depression are, in this case, consequences of the cancer diagnosis,⁷ although the exact mechanisms are not understood. Being anxious or depressed may prompt further changes in behavior, such as social withdrawal. So there may be social consequences to the diagnosis as well. This example is designed to lay out some of the complexity of the biopsychosocial model applied to mental health and mental illness.

Biological Influences on Mental Health and Mental Illness

There are far-reaching biological and physical influences on mental health and mental illness. The major categories are genes, infections, physical trauma, nutrition, hormones, and toxins (e.g., lead). Examples have been noted throughout Chapter 1 and earlier in this chapter. This section focuses on the first two categories—genes and infections—for these are among the most exciting and intensive areas of research relating to biological influences on mental health and mental illness

The Genetics of Behavior and Mental Illness

That genes influence behavior, normal and abnormal, has long been established (Plomin et al., 1997). Genes influence behavior across the animal spectrum, from the lowly fruitfly all the way to humans. Sorting out which genes are involved and determining how they influence behavior present the greatest challenge. Research suggests that many mental disorders arise in part from defects not in single genes, but in multiple genes. However, none of the genes has yet been pinpointed for common mental disorders (National Institute of Mental Health [NIMH], 1998).

The human genome contains approximately 80,000 genes that occupy approximately 5 percent of the DNA sequences of the human genome. By the spring of 2000, the human genome project will have provided an initial rough draft version of the entire sequence of the human genome, and in the ensuing years, gaps in the sequence will be closed, errors will be corrected, and the precise boundaries of genes will be identified.

In parallel, clinical medicine is studying the aggregation of human disease in families. This effort includes the study of mental illness, most notably schizophrenia, bipolar disorder (manic depressive illness), early onset depression, autism, attention-deficit/hyperactivity disorder, anorexia nervosa, panic disorder, and a number of other mental disorders (NIMH, 1998). From studying how these disorders run in families, and from initial molecular analyses of the genomes of these families, we have learned that heredity—that is, genes—plays a role in the transmission of vulnerability of all the aforementioned disorders from generation to generation.

But we have also learned that the transmission of risk is not simple. Certain human diseases such as Huntington’s disease and cystic fibrosis result from the transmission of a mutation—that is, a deleteriously altered gene sequence—at one location in the human genome. In these diseases, a single mutation has everything to say about whether one will get the illness. The transmission of a trait due to a single gene in the human genome is called Mendelian transmission, after the Austrian monk, Gregor Mendel, who was the first to develop principles of modern genetics and who studied traits due to single genes. When a single gene determines the presence or absence of a disease or other trait, genes are rather easy to discover on the basis of modern methods. Indeed, for almost all Mendelian disorders across medicine that affect more than a few people, the genes already have been identified.

In contrast to Mendelian disorders, to our knowledge, all mental illnesses and all normal variants of behavior are genetically complex. What this means is that no single gene or even a combination of genes dictates whether someone will have an illness or a particular behavioral trait. Rather, mental illness appears to result from the interaction of multiple genes that confer risk, and this risk is converted into illness by the interaction of genes with environmental factors. The implications for science are, first, that no gene is equivalent to fate for mental illness. This gives us hope that modifiable environmental risk factors can eventually be identified and become targets for prevention efforts. In addition, we recognize that genes, while significant in their aggregate contribution to risk, may each contribute only a small increment, and, therefore, will be difficult to discover. As a result, however, of the Human Genome Project, we will know the sequence of each human gene and the common variants for each gene throughout the human race. With this information, combined with modern technologies, we will in the coming years identify genes that confer risk of specific mental illnesses.

This information will be of the highest importance for several reasons. First, genes are the blueprints of cells. The products of genes, proteins, work together in pathways or in building cellular structures, so that finding variants within genes will suggest pathways that can be targets of opportunity for the development of new therapeutic interventions. Genes will also be important clues to what goes wrong in the brain when a disease occurs. For example, once we know that a certain gene is involved in risk of a particular mental illness such as schizophrenia or autism, we can ask at what time during the development of the brain that particular gene is active and in which cells and circuits the gene is expressed. This will give us clues to critical times for intervention in a disease process and information about what it is that goes wrong. Finally, genes will provide tools for those scientists who are searching for environmental risk factors. Information from genetics will tell us at what age environmental cofactors in risk must be active, and genes will help us identify homogeneous populations for studies of treatment and of prevention.

Heritability refers to how much genetics contributes to the variation of a disease or trait in a population at a given point in time (Plomin et al., 1997). Once a disorder is established as running in families, the next step is to determine its heritability (see below), then its mode of transmission, and, lastly, its location through genetic mapping (Lombroso et al., 1994).

One powerful method for estimating heritability is through twin studies.⁸ Twin studies often compare the frequency with which identical versus fraternal twins display a disorder. Since identical twins are from the same fertilized egg, they share the exact genetic inheritance. Fraternal twins are from separate eggs and thereby share only 50 percent of their genetic inheritance. If a disorder is heritable, identical twins should have a higher rate of concordance—the expression of the trait by both members of a twin pair—than fraternal twins. Such studies, however, do not furnish information about which or how many genes are involved. They just can be used to estimate heritability. For example, the heritability of bipolar disorder, according to the most rigorous twin study, is about 59 percent, although other estimates vary (NIMH, 1998). The heritability of schizophrenia is estimated, on the basis of twin studies, at a somewhat higher level (NIMH, 1998).

Even with a high level of heritability, however, it is essential to point out that environmental factors (e.g., psychosocial environment, nutrition, health care access) can play a significant role in the severity and course of a disorder.

Another point is that environmental factors may even protect against the disorder developing in the first place. Even with the relatively high heritability of schizophrenia, the median concordance rate among identical twins is 46 percent⁹ (NIMH, 1998), meaning that in over half of the cases, the second twin does not manifest schizophrenia even though he or she has the same genes as the affected twin. This implies that environmental factors exert a significant role in the onset of schizophrenia.

Infectious Influences

It has been known since the early part of the 20th century that infectious agents can penetrate into the brain where they can cause mental disorders. A highly common mental disorder of unknown etiology at the turn of the century, termed “general paresis,” turned out to be a late manifestation of syphilis. The sexually transmitted infectious agent—Treponema pallidum—first caused symptoms in reproductive organs and then, sometimes years later, migrated to the brain where it led to neurosyphilis. Neurosyphilis was manifest by neurological deterioration (including psychosis), paralysis, and later death. With the wide availability of penicillin after World War II, neurosyphilis was virtually eliminated (Barondes, 1993).

Neurosyphilis may be thought of as a disease of the past (at least in the developed world), but dementia associated with infection by the human immunodeficiency virus (HIV) is certainly not. HIV-associated dementia continues to encumber HIV-infected individuals worldwide. HIV infection penetrates into the brain, producing a range of progressive cognitive and behavioral impairments. Early symptoms include impaired memory and concentration, psychomotor slowing, and apathy. Later symptoms, usually appearing years after infection, include global impairments marked by mutism, incontinence, and paraplegia (Navia et al., 1986). The prevalence of HIV-associated dementia varies, with estimates ranging from 15 percent to 44 percent of patients with HIV infection (Grant et al., 1987; McArthur et al., 1993). The high end of this estimate includes patients with subtle neuropsychological abnormalities. What is remarkable about HIV-associated dementia is that it appears to be caused not by direct infection of neurons, but by infection of immune cells known as macrophages that enter the brain from the blood. The macrophages indirectly cause dysfunction and death in nearby neurons by releasing soluble toxins (Epstein & Gendelman, 1993).

Besides HIV-associated dementia and neurosyphilis, other mental disorders are caused by infectious agents. They include herpes simplex encephalitis, measles encephalomyelitis, rabies encephalitis, chronic meningitis, and subacute sclerosing panencephalitis (Kaplan & Sadock, 1998). More recently, research has uncovered an infectious etiology to one form of obsessive-compulsive disorder, as explained below.

PANDAS

In the late 1980s, it was discovered that some children with obsessive-compulsive disorder (OCD) experienced a sudden onset of symptoms soon after a streptococcal pharyngitis (Garvey et al., 1998). The symptoms were classic for OCD—concerns about contamination, spitting compulsions, and extremely excessive hoarding—but the abrupt onset was unusual. Further study of these children led to the identification of a new classification of OCD called PANDAS. This acronym stands for pediatric autoimmune neuropsychiatric disorders associated with streptococcal infection. PANDAS are distinct from classic cases of OCD because of their episodic clinical course marked by sudden symptom exacerbation linked to streptococcal infection, among other unique features. The exacerbation of symptoms is correlated with a rise in levels of antibodies that the child produces to fight the strep infection. Consequently, researchers proposed that PANDAS are caused by antibodies against the strep infection that also manage to attack the basal ganglia region of the child’s brain (Garvey et al., 1998). In other words, the strep infection triggers the child’s immune system to develop antibodies, which, in turn, may attack the child’s brain, leading to obsessive and compulsive behaviors. Under this proposal, the strep infection does not directly induce the condition; rather, it may do so indirectly by triggering antibody formation. How the antibodies are so damaging to a discrete region of the child’s brain and how this attack ignites OCD-like symptoms are two of the fundamental questions guiding research.

Psychosocial Influences on Mental Health and Mental Illness

This chapter thus far has highlighted some of the psychosocial influences on mental health and mental illness. Stressful life events, affect (mood and level of arousal), personality, and gender are prominent psychological influences. Social influences include parents, socioeconomic status, racial, cultural, and religious background, and interpersonal relationships. These psychosocial influences, taken individually or together, are integrated into many chapters of this report in discussions of epidemiology, etiology, risk factors, barriers to treatment, and facilitators to recovery.

Since these psychosocial influences are familiar to the general reader, detailed description of each is beyond the scope of this section (with the exception of cultural influences, which are discussed in the Overview of Cultural Diversity and Mental Health Services section). Instead, this section summarizes the sweeping theories of individual behavior and personality that inspired a vast body of psychosocial research: psychodynamic theories, behaviorism, and social learning theories. The therapeutic strategies that arose from these theories, and modifications necessary to make them relevant to the changing demography of the U.S. population, are discussed in a later section, Overview of Treatment.

Psychodynamic Theories

Psychodynamic theories of personality assert that behavior is the product of underlying conflicts over which people often have scant awareness. Sigmund Freud (1856–1939) was the towering proponent of psychoanalytic theory, the first of the 20th-century psychodynamic theories. Many of Freud’s followers pioneered their own psychodynamic theories, but this section covers only psychoanalytic theory. A brief discussion of Freud’s work contributes to a historical perspective of mental health theory and treatment approaches.

Freud’s theory of psychoanalysis holds two major assumptions: (1) that much of mental life is unconscious (i.e., outside awareness), and (2) that past experiences, especially in early childhood, shape how a person feels and behaves throughout life (Brenner, 1978).

Freud’s structural model of personality divides the personality into three parts—the id, the ego, and the superego. The id is the unconscious part that is the cauldron of raw drives, such as for sex or aggression. The ego, which has conscious and unconscious elements, is the rational and reasonable part of personality. Its role is to maintain contact with the outside world in order to help keep the individual in touch with society. As such, the ego mediates between the conflicting tendencies of the id and the superego. The latter is a person’s conscience that develops early in life and is learned from parents, teachers, and others. Like the ego, the superego has conscious and unconscious elements (Brenner, 1978).

When all three parts of the personality are in dynamic equilibrium, the individual is thought to be mentally healthy. However, according to psychoanalytic theory, if the ego is unable to mediate between the id and the superego, an imbalance would occur in the form of psychological distress and symptoms of mental disorders. Psychoanalytic theory views symptoms as important only in terms of expression of underlying conflicts between the parts of personality. The theory holds that the conflicts must be understood by the individual with the aid of the psychoanalyst who would help the person unearth the secrets of the unconscious. This was the basis for psychoanalysis as a form of treatment, as explained later in this chapter.

Behaviorism and Social Learning Theory

Behaviorism (also called learning theory) posits that personality is the sum of an individual’s observable responses to the outside world (Feldman, 1997). As charted by J. B. Watson and B. F. Skinner in the early part of the 20th century, behaviorism stands at loggerheads with psychodynamic theories, which strive to understand underlying conflicts. Behaviorism rejects the existence of underlying conflicts and an unconscious. Rather, it focuses on observable, overt behaviors that are learned from the environment (Kazdin, 1996, 1997). Its application to treatment of mental problems, which is discussed later, is known as behavior modification.

Learning is seen as behavior change molded by experience. Learning is accomplished largely through either classical or operant conditioning. Classical conditioning is grounded in the research of Ivan Pavlov, a Russian physiologist. It explains why some people react to formerly neutral stimuli in their environment, stimuli that previously would not have elicited a reaction. Pavlov’s dogs, for example, learned to salivate merely at the sound of the bell, without any food in sight. Originally, the sound of the bell would not have elicited salvation. But by repeatedly pairing the sight of the food (which elicits salvation on its own) with the sound of the bell, Pavlov taught the dogs to salivate just to the sound of the bell by itself.

Operant conditioning, a process described and coined by B. F. Skinner, is a form of learning in which a voluntary response is strengthened or attenuated, depending on its association with positive or negative consequences (Feldman, 1997). The strengthening of responses occurs by positive reinforcement, such as food, pleasurable activities, and attention from others. The attenuation or discontinuation of responses occurs by negative reinforcement in the form of removal of a pleasurable stimulus. Thus, human behavior is shaped in a trial and error way through positive and negative reinforcement, without any reference to inner conflicts or perceptions. What goes on inside the individual is irrelevant, for humans are equated with“black boxes.” Mental disorders represented maladaptive behaviors that were learned. They could be unlearned through behavior modification (behavior therapy) (Kazdin, 1996, 1997).

The movement beyond behaviorism was spearheaded by Albert Bandura (1969, 1977), the originator of social learning theory (also known as social cognitive theory). Social learning theory has its roots in behaviorism, but it departs in a significant way. While acknowledging classical and operant conditioning, social learning theory places far greater emphasis on a different type of learning, particularly observational learning. Observational learning occurs through selectively observing the behavior of another person, a model. When the behavior of the model is rewarded, children are more likely to imitate the behavior. For example, a child who observes another child receiving candy for a particular behavior is more likely to carry out similar behaviors. Social learning theory asserts that people’s cognitions—their views, perceptions, and expectations toward their environment—affect what they learn. Rather than being passively conditioned by the environment, as behaviorism proposed, humans take a more active role in deciding what to learn as a result of cognitive processing. Social learning theory gave rise to cognitive-behavioral therapy, a mode of treatment described later in this chapter and throughout this report.

The Integrative Science of Mental Illness and Health

Progress in understanding depression and schizophrenia offers exciting examples of how findings from different disciplines of the mental health field have many common threads (Andreasen, 1997). Despite the differences in terminology and methodology, the results from different disciplines have converged to paint a vivid picture of the nature of the fundamental defects and the regions of the brain that underlie these defects. Even in the case of depression and schizophrenia, there is much to be uncovered about etiology, yet the mental health field is seen as poised“to use the power of multiple disciplines.” The disciplines are urged to link together the study of the mind and the brain in the search for understanding mental health and mental illness (Andreasen, 1997).

This linkage already has been cemented between cognitive psychology, behavioral neurology, computer science, and neuroscience. These disciplines have knit together the field of “cognitive neuroscience” (Kosslyn & Shin, 1992). This new and joint discipline has carved out its own professional society, journals (Waldrop, 1993), and textbooks (Gazzaniga et al., 1998). There is movement toward integration of other disciplines within the field. To promote linkages between psychiatry and the neurosciences, neuroscientist Eric R. Kandel has furnished a novel approach. His essay,“A New Intellectual Framework for Psychiatry,” supplies a set of biological principles to forge a rapprochement—conceptual as well as practical—between the two disciplines (Kandel, 1998). Integrated approaches are seen as vital to tackle the monumental complexity of mental function.

⁵ Chapter 4 contains a fuller discussion of the relationship between stress and depression. In common parlance, stress refers either to the stressful event or to the individual’s response to the event. However, mental health professionals distinguish the two by referring to the external events as the “stressor” (or stressful life event) and to the individual’s response as the “stress response.”

⁶ Other types of information used to establish cause and effect relationships are the strength and consistency of the association, time sequence information, dose-response relationships, and disappearance of the effect when the cause is removed.

⁷ Anxiety and depression may in some cases be caused by hormonal changes related to the tumor itself.

⁸ Establishing that a disorder runs in families could suggest environmental and/or genetic influences because families share genes and environment. Comparing identical versus fraternal twins assumes that their shared environments are about equal, thereby providing insight about genetic influences. Such comparisons are further enhanced by studies of twins (identical vs. fraternal) separated at birth and adopted by different families.

⁹ The median concordance rate for identical twins is only 14 percent (NIMH, 1998).