Of the biogenic amines, norepinephrine and serotonin are the two neurotransmitters most implicated in the pathophysiology of mood disorders.

Acetylcholine (ACh) is found in neurons that are distributed diffusely throughout the cerebral cortex. Cholinergic neurons have reciprocal or interactive relationships with all three monoamine systems. Abnormal levels of choline, which is a precursor to ACh, have been found at autopsy in the brains of some depressed patients, perhaps reflecting abnormalities in cell phospholipid composition. Cholinergic agonist and antagonist drugs have differential clinical effects on depression and mania. Agonists can produce lethargy, anergia, and psychomotor retardation in healthy subjects, can exacerbate symptoms in depression, and can reduce symptoms in mania. These effects generally are not sufficiently robust to have clinical applications, and adverse effects are problematic. In an animal model of depression, strains of mice that are super- or subsensitive to cholinergic agonists have been found to be susceptible or more resistant to developing learned helplessness (discussed later). Cholinergic agonists can induce changes in hypothalamic-pituitary-adrenal (HPA) activity and sleep that mimic those associated with severe depression. Some patients with mood disorders in remission, as well as their never-ill first-degree relatives, have a trait-like increase in sensitivity to cholinergic agonists.

γ-Aminobutyric acid (GABA) has an inhibitory effect on ascending monoamine pathways, particularly the mesocortical and mesolimbic systems. Reductions have been observed in plasma, CSF, and brain GABA levels in depression. Animal studies have also found that chronic stress can reduce and eventually deplete GABA levels. By contrast, GABA receptors are upregulated by antidepressants, and some GABAergic medications have weak antidepressant effects.

The amino acids glutamate and glycine are the major excitatory and inhibitory neurotransmitters in the CNS. Glutamate and glycine bind to sites associated with the N-methyl-D-aspartate (NMDA) receptor, and an excess of glutamatergic stimulation can cause neurotoxic effects. Of importance, a high concentration of NMDA receptors exists in the hippocampus. Glutamate, thus, may work in conjunction with hypercortisolemia to mediate the deleterious neurocognitive effects of severe recurrent depression. Emerging evidence suggests that drugs that antagonize NMDA receptors have antidepressant effects.

The binding of a neurotransmitter and a postsynaptic receptor triggers a cascade of membrane-bound and intracellular processes mediated by second messenger systems. Receptors on cell membranes interact with the intracellular environment via guanine nucleotide-binding proteins (G proteins). The G proteins, in turn, connect to various intracellular enzymes (e.g., adenylate cyclase, phospholipase C, and phosphodiesterase) that regulate utilization of energy and formation of second messengers, such as cyclic nucleotide (e.g., cyclic adenosine monophosphate [cAMP] and cyclic guanosine monophosphate), as well as phosphatidylinositols (e.g., inositol triphosphate and diacylglycerol) and calcium-calmodulin. Second messengers regulate the function of neuronal membrane ion channels. Increasing evidence also indicates that mood-stabilizing drugs act on G proteins or other second messengers.

Lasting alterations in neuroendocrine and behavioral responses can result from severe early stress. Animal studies indicate that even transient periods of maternal deprivation can alter subsequent responses to stress. Activity of the gene coding for the neurokinin brain-derived neurotrophic growth factor is decreased after chronic stress, as is the process of neurogenesis. Protracted stress thus can induce changes in the functional status of neurons and, eventually, cell death. Recent studies in depressed humans indicate that a history of early trauma is associated with increased HPA activity accompanied by structural changes (i.e., atrophy or decreased volume) in the cerebral cortex.

Depression is associated with a premature loss of deep (slow wave) sleep and an increase in nocturnal arousal. The latter is reflected by four types of disturbance: (1) an increase in nocturnal awakenings, (2) a reduction in total sleep time, (3) increased phasic rapid eye movement (REM) sleep, and (4) increased core body temperature. The combination of increased REM drive and decreased slow wave sleep results in a significant reduction in the first period of non-REM (NREM) sleep, a phenomenon referred to as reduced REM latency. Reduced REM latency and deficits of slow wave sleep typically persist after recovery of a depressive episode. Blunted secretion of GH after sleep onset is associated with decreased slow wave sleep and shows similar state-independent or trait-like behavior. The combination of reduced REM latency, increased REM density, and decreased sleep maintenance identifies approximately 40 percent of depressed outpatients and 80 percent of depressed inpatients. False-negative findings are commonly seen in younger, hypersomnolent patients, who may actually experience an increase in slow wave sleep during episodes of depression. Approximately 10 percent of otherwise healthy individuals have abnormal sleep profiles, and, as with dexamethasone nonsuppression, false-positive cases are not uncommonly seen in other psychiatric disorders.

Patients manifesting a characteristically abnormal sleep profile have been found to be less responsive to psychotherapy and to have a greater risk of relapse or recurrence and may benefit preferentially from pharmacotherapy.

Depressive disorders are associated with several immunological abnormalities, including decreased lymphocyte proliferation in response to mitogens and other forms of impaired cellular immunity. These lymphocytes produce neuromodulators, such as corticotropin-releasing factor, and cytokines, peptides known as interleukins. There appears to be an association with clinical severity, hypercortisolism, and immune dysfunction, and the cytokine interleukin-1 may induce gene activity for glucocorticoid synthesis.

Computed axial tomography and magnetic resonance imaging scans have permitted the use of sensitive, noninvasive methods to assess the living brain, including cortical and subcortical tracts, as well as white matter lesions. The most consistent abnormality observed in the depressive disorders is increased frequency of abnormal hyperintensities in subcortical regions, such as periventricular regions, the basal ganglia, and the thalamus. More common in bipolar I disorder and among the elderly, these hyperintensities appear to reflect the deleterious neurodegenerative effects of recurrent affective episodes. Ventricular enlargement, cortical atrophy, and sulcal widening also have been reported in some studies. Some depressed patients also may have reduced hippocampal or caudate nucleus volumes, or both, suggesting more focal defects in relevant neurobehavioral systems. Diffuse and focal areas of atrophy have been associated with increased illness severity, bipolarity, and increased cortisol levels.

The most widely replicated positron emission tomography finding in depression is decreased anterior brain metabolism, which is generally more pronounced on the left side. From a different vantage point, depression may be associated with a relative increase in nondominant hemispheric activity. Furthermore, a reversal of hypofrontality occurs after shifts from depression into hypomania, such that greater left hemisphere reductions are seen in depression compared with greater right hemisphere reductions in mania. Other studies have observed more specific reductions of reduced cerebral blood flow or metabolism, or both, in the dopaminergically innervated tracts of the mesocortical and mesolimbic systems in depression. Again, evidence suggests that antidepressants at least partially normalize these changes.

In addition to a global reduction of anterior cerebral metabolism, increased glucose metabolism has been observed in several limbic regions, particularly among patients with relatively severe recurrent depression and a family history of mood disorder. During episodes of depression, increased glucose metabolism is correlated with intrusive ruminations.

Both the symptoms of mood disorders and biological research findings support the hypothesis that mood disorders involve pathology of the brain. Modern affective neuroscience focuses on the importance of four brain regions in the regulation of normal emotions: the prefrontal cortex (PFC), the anterior cingulate, the hippocampus, and the amygdala. The PFC is viewed as the structure that holds representations of goals and appropriate responses to obtain these goals. Such activities are particularly important when multiple, conflicting behavioral responses are possible or when it is necessary to override affective arousal. Evidence indicates some hemispherical specialization in PFC function. For example, left-sided activation of regions of the PFC is more involved in goal-directed or appetitive behaviors, whereas regions of the right PFC are implicated in avoidance behaviors and inhibition of appetitive pursuits. Subregions in the PFC appear to localize representations of behaviors related to reward and punishment.

The anterior cingulate cortex (ACC) is thought to serve as the point of integration of attentional and emotional inputs. Two subdivisions have been identified: an affective subdivision in the rostral and ventral regions of the ACC and a cognitive subdivision involving the dorsal ACC. The former subdivision shares extensive connections with other limbic regions, and the latter interacts more with the PFC and other cortical regions. It is proposed that activation of the ACC facilitates control of emotional arousal, particularly when goal attainment has been thwarted or when novel problems have been encountered.

The hippocampus is most clearly involved in various forms of learning and memory, including fear conditioning, as well as inhibitory regulation of HPA-axis activity. Emotional or contextual learning appears to involve a direct connection between the hippocampus and the amygdala.

The amygdala appears to be a crucial way station for processing novel stimuli of emotional significance and coordinating or organizing cortical responses. Located just above the hippocampi bilaterally, the amygdala has long been viewed as the heart of the limbic system. Although most research has focused on the role of the amygdala in responding to fearful or painful stimuli, it may be ambiguity or novelty, rather than the aversive nature of the stimulus per se, that brings the amygdala on line.

Family studies address the question of whether a disorder is familial. More specifically, is the rate of illness in the family members of someone with the disorder greater than that of the general population? Family data indicate that if one parent has a mood disorder, a child will have a risk of between 10 and 25 percent for mood disorder. If both parents are affected, this risk roughly doubles. The more members of the family who are affected, the greater the risk is to a child. The risk is greater if the affected family members are first-degree relatives rather than more distant relatives. A family history of bipolar disorder conveys a greater risk for mood disorders in general and, specifically, a much greater risk for bipolar disorder. Unipolar disorder is typically the most common form of mood disorder in families of bipolar probands. This familial overlap suggests some degree of common genetic underpinning between these two forms of mood disorder. The presence of more-severe illness in the family also conveys a greater risk.

Adoption studies provide an alternative approach to separating genetic and environmental factors in familial transmission. Only a limited number of such studies have been reported, and their results have been mixed. One large study found a threefold increase in the rate of bipolar disorder and a twofold increase in unipolar disorder in the biological relatives of bipolar probands. Similarly, in a Danish sample, a threefold increase in the rate of unipolar disorder and a sixfold increase in the rate of completed suicide in the biological relatives of affectively ill probands were reported. Other studies, however, have been less convincing and have found no difference in the rates of mood disorders.

Twin studies provide the most powerful approach to separating genetic from environmental factors, or “nature” from “nurture.” The twin data provide compelling evidence that genes explain only 50 to 70 percent of the etiology of mood disorders. Environment or other nonheritable factors must explain the remainder. Therefore, it is a predisposition or susceptibility to disease that is inherited. Considering unipolar and bipolar disorders together, these studies find a concordance rate for mood disorder in monozygotic twins of 70 to 90 percent compared with a rate in same-sex dizygotic twins of 16 to 35 percent. This is the most compelling data for the role of genetic factors in mood disorders.

Linkage Studies. Deoxyribonucleic acid (DNA) markers are segments of DNA of known chromosomal location, which are highly variable among individuals. They are used to track the segregation of specific chromosomal regions within families affected with a disorder. When a marker is identified with disease in families, the disease is said to be genetically linked. Chromosomes 18q and 22q are the two regions with strongest evidence for linkage to bipolar disorder. Several linkage studies have found evidence for the involvement of specific genes in clinical subtypes. For example, the linkage evidence on 18q has been shown to be derived largely from bipolar II-bipolar II sibling pairs and from families in which the probands had panic symptoms.

Gene-mapping studies of unipolar depression have found very strong evidence of linkage to the locus for cAMP response element-binding protein (CREB1) on chromosome 2. Eighteen other genomic regions were found to be linked; some of these displayed interactions with the CREB1 locus. Another study has reported evidence for a gene-environment interaction in the development of major depression. Subjects who underwent adverse life events were shown, in general, to be at an increased risk for depression. Of such subjects, however, those with a variant in the serotonin transporter gene showed the greatest increase in risk. This is one of the first reports of a specific gene-environment interaction in a psychiatric disorder.

A long-standing clinical observation is that stressful life events more often precede first, rather than subsequent, episodes of mood disorders. This association has been reported for both patients with major depressive disorder and patients with bipolar I disorder. One theory proposed to explain this observation is that the stress accompanying the first episode results in long-lasting changes in the brain’s biology. These long-lasting changes may alter the functional states of various neurotransmitter and intraneuronal signaling systems, changes that may even include the loss of neurons and an excessive reduction in synaptic contacts. As a result, a person has a high risk of undergoing subsequent episodes of a mood disorder, even without an external stressor.

Some clinicians believe that life events play the primary or principal role in depression; others suggest that life events have only a limited role in the onset and timing of depression. The most compelling data indicate that the life event most often associated with development of depression is losing a parent before age 11 years. The environmental stressor most often associated with the onset of an episode of depression is the loss of a spouse. Another risk factor is unemployment; persons out of work are three times more likely to report symptoms of an episode of major depression than those who are employed.

No single personality trait or type uniquely predisposes a person to depression; all humans, of whatever personality pattern, can and do become depressed under appropriate circumstances. Persons with certain personality disorders—obsessive-compulsive, histrionic, and borderline—may be at greater risk for depression than persons with antisocial or paranoid personality disorder. The latter can use projection and other externalizing defense mechanisms to protect themselves from their inner rage. No evidence indicates that any particular personality disorder is associated with later development of bipolar I disorder; however, patients with dysthymic disorder and cyclothymic disorder are at risk of later developing major depression or bipolar I disorder.

Recent stressful events are the most powerful predictors of the onset of a depressive episode. From a psychodynamic perspective, the clinician is always interested in the meaning of the stressor. Research has demonstrated that stressors that the patient experiences as reflecting negatively on his or her self-esteem are
more likely to produce depression. Moreover, what may seem to be a relatively mild stressor to outsiders may be devastating to the patient because of particular idiosyncratic meanings attached to the event.

The psychodynamic understanding of depression defined by Sigmund Freud and expanded by Karl Abraham is known as the classic view of depression. That theory involves four key points: (1) disturbances in the infant-mother relationship during the oral phase (the first 10 to 18 months of life) predispose to subsequent vulnerability to depression; (2) depression can be linked to real or imagined object loss; (3) introjection of the departed objects is a defense mechanism invoked to deal with the distress connected with the object’s loss; and (4) because the lost object is regarded with a mixture of love and hate, feelings of anger are directed inward at the self.

Melanie Klein understood depression as involving the expression of aggression toward loved ones, much as Freud did. Edward Bibring regarded depression as a phenomenon that sets in when a person becomes aware of the discrepancy between extraordinarily high ideals and the inability to meet those goals. Edith Jacobson saw the state of depression as similar to the situation of a powerless, helpless child victimized by a tormenting parent. Silvano Arieti observed that many depressed people have lived their lives for someone else rather than for themselves. He referred to the person for whom depressed patients live as the dominant other, which may be a principle, an ideal, or an institution, as well as an individual. Depression sets in when patients realize that the person or ideal for which they have been living is never going to respond in a manner that will meet their expectations. Heinz Kohut’s conceptualization of depression, derived from his self-psychological theory, rests on the assumption that the developing self has specific needs that must be met by parents to give the child a positive sense of self-esteem and self-cohesion. When others do not meet these needs, there is a massive loss of self-esteem that presents as depression. John Bowlby believed that damaged early attachments and traumatic separation in childhood predispose to depression. Adult losses are said to revive the traumatic childhood loss and so precipitate adult depressive episodes.