Anxiety and depression are among the most prevalent psychiatric disorders in childhood [10.5% of school-aged children are affected in Ontario (1)]. Many studies emphasize that the long-term morbidity and disability associated with these disorders is severe [reviewed in (2)]. In adolescents, depression is also associated with increased mortality due to suicide (1).

Until recently, childhood anxiety and depression were underreported. There is some debate as to whether the apparent increase in these conditions in recent years is due to stresses in children’s lives or is an artifact of better reporting. Whichever is the case, many affected children still experience lengthy delays in seeking and obtaining treatment. The large Ontario epidemiologic study concluded that up to 80% of children with these disorders are never treated (1). Moreover, a substantial number of anxious children experience comorbid depression, and the risk of this comorbidity increases with age and lack of intervention (3). In contrast to children with either disorder alone, children with comorbid anxiety and depression often suffer sequelae well into adulthood and are relatively less responsive to treatment (4). These findings underscore the need to identify anxious and depressed children at an early age, to improve their functioning, and to reduce the risk of comorbidity and long-term impairment.

Sex differences in rates of anxiety and depression are not consistently found in childhood, but there is a clear female preponderance (about 2:1 for depression) in both conditions by adolescence [see (5)]. Possible reasons for this shift in sex distribution are discussed below (see also Chapter 12, on depression and anxiety in adulthood).

Measurement difficulties may contribute to underidentification and treatment delay in these conditions. Several well-validated self-report inventories exist for anxiety and depression [for example, the Multidimensional Anxiety Scale for Children (6); Children’s Depression Inventory (7)]. However, correspondence among informants is consistently fair to poor [reviewed in (8)], with parents tending to report more overt, behavioral symptoms and children reporting more distressed feeling states. In some cases, a child may fail to report symptoms despite evidence of disorder. Alternatively, the child may recognize symptoms although adults in his or her environment do not and hence referral for assessment and treatment is unlikely to occur. Therefore, in clinical practice, symptoms that either the parent or the child endorse merit investigation.

Symptom inventories are also limited by low specificity because many children with noninternalizing types of psychopathology have elevated scores. Such inventories are best used as part of a more thorough diagnostic assessment that provides an estimate of severity and (when repeated) a means of evaluating progress objectively.

Anxiety and depression require somewhat different thresholds for concern. Certain fears are developmentally normative [reviewed in (9)], and mild anxiety enhances academic performance on some tasks. Avoidance of risk can also be adaptive in certain circumstances (for example, reducing risk of accident-related injuries). Therefore, anxiety is considered pathologic only if it is very distressing
or results in substantial impairment. By contrast, the presence of depressive symptoms, even at subclinical levels, always merits concern. Large-scale longitudinal studies of children with elevated depression scores on self-report found a threefold increased risk for subsequent depression compared to other children (10,11).

Twin and adoption studies have shown that genetic factors account for at least 50% of the variance in the transmission of mood disorders, although specific genes are only now being identified (12). Children of depressed parents have a threefold lifetime risk of major depression and are also more vulnerable to anxiety disorders. Therefore, some authors have suggested that what is inherited is a general tendency to internalize problems and that environmental factors influence which symptoms are manifested.

The Virginia Twin Study has yielded some interesting findings with respect to the relationship between anxiety and depression in girls (13). Using over 600 female twin pairs (about two thirds monozygotic and one third dizygotic), the authors found three distinct patterns. Early overanxious disorder (OAD) and simple phobias were linked to specific genetic vulnerability to depression after age 14; shared environmental factors influenced symptoms of depression before age 14 and later separation anxiety and simple phobias; and shared environmental influences on depression after age 14 also increased vulnerability to concurrent OAD and separation anxiety. The association between early OAD and later depression is consistent with models suggesting a temporal progression from anxiety to depression. Interestingly, these findings suggest that the progression may be based, at least in part, on a common genetic risk rather than a depressive response to untreated anxiety symptoms. Other studies have since confirmed shared environmental influences as more salient in prepubertal depression, with greater heritability in adolescent depression (14). Genetic susceptibility to adolescent depression in response to environmental adverse events was also demonstrated in this sample.

Family and twin studies have both found a definite familial contribution to childhood anxiety disorders [reviewed in (2)]. Anxiety disorder rates were elevated in children of parents with anxiety disorders or mixed anxiety-depression. Conversely, parents of children with school refusal were found to have elevated rates of anxiety and depressive disorders. While several twin studies have established a genetic contribution to childhood anxiety, further studies will clarify the degree of genetic versus environmental contribution to specific anxiety disorders. In a recent twin study, for example, Topolski and others (15) found that shared environmental effects played a moderate role in separation anxiety disorder but not in OAD. In contrast, genetic influences appeared to predominate in OAD. Specific genetic loci are now being researched.

The constitutional factor examined most extensively in relation to childhood anxiety is a trait termed “behavioral inhibition” [reviewed in (16)]. Behavioral inhibition is an aspect of temperament measurable in the laboratory, characterized by a tendency to restrict exploration and avoid novelty. Human and primate studies have confirmed the heritability of this trait, and its physiologic basis. Inhibited children show evidence of chronically high sympathetic arousal, leading to the hypothesis that inhibition occurs when there is a reduced threshold to arousal in the amygdala, a part of the limbic system.

Many children “outgrow” behavioral inhibition, probably because of their parents’ encouraging them to enter new situations, which results in desensitization. There may be sex differences, however, in the degree of encouragement of exploratory behavior parents offer. In many cultures, girls are expected to be more reticent than boys, resulting in inhibited behavior being socially sanctioned and thus likely to persist. Persistent inhibition, in turn, has been linked prospectively to an increased risk of subsequent anxiety disorders in middle childhood and specifically to social phobia in adolescence.

Perhaps because they typically have a later onset, less study has been devoted to temperamental antecedents of mood disorders.

Cognitive biases have been implicated in childhood depression [reviewed in (17)]. For example, longitudinal studies have shown that children with a negative cognitive style experience prolonged dysphoric mood when exposed to stresses such as poor grades or peer rejection. Such dysphoric mood may predispose to depression. Negative cognitive style and self-perceptions are also found during depressive episodes, but it is unclear whether these are part of the depressed state or represent long-term traits. Modeling, critical or rejecting familial relationships, and stressful life events all have been proposed as contributing to negative cognitions in depressed children.

Cognitive biases, in particular attentional and interpretational biases, are also thought to play a role in many anxiety disorders, with relative specificity depending on the type of bias [reviewed in (18)]. For example, when doing a timed color-naming task of both neutral and threatening words, spider-phobic children had slowed color-naming times when shown spider-related words (e.g., web) but not neutral words (e.g., fly). Anxious children were also faster to react to a probe preceded by a threatening rather than a neutral word on dot-probe tasks, showing selective attention to threat cues.

Interpretation bias has been evidenced in studies comparing anxious versus nonanxious children’s interpretation of homophones (e.g., dye or die) and ambiguous stories. Results indicated that anxious children were more likely to interpret ambiguous situations in a threatening way. Parent-child interactions can, however, influence children’s tendency to interpret ambiguous material as threatening.

Dichotic listening tasks that examined cerebral laterality have also found differences between anxious and depressed children relative to normal controls. My colleagues and I (19) found enhanced perception of emotional tone of voice in anxious children compared to normal children, regardless of ear or hemisphere of presentation. Pine and colleagues (20) found that depressed adolescents showed an increased right ear/left hemisphere advantage for fused words.

In summary, there is emerging evidence that cognitive biases in attention and perception likely have some physiologic basis, but interpretation biases may also be subject to the influence of parental modeling and other parent-child interactions. None of the cognitive bias studies found gender differences, but parental expectations of girls might still play a role in interpretation bias.

Following upon work in adults, several hormones have been studied in relation to childhood depression. [reviewed in (12)]. Because most studies are cross-sectional, it is often unclear whether these constitute vulnerability factors or markers of the depressed state. Blunted response of growth hormones to insulin-induced hypoglycemia has been found to persist upon remission, suggesting that this may be more of a trait marker. Findings regarding growth hormone without stimulation are equivocal, though. Findings in children are also inconsistent for the dexamethasone-suppression test, a common marker for depression in adults. Lower cortisol levels after infusion of L-5-hydroxytryptophan have been found in children with early-onset major depression. Depressed girls have been found to have elevated prolactin levels compared to girls who are not depressed. None of these findings is specific enough to constitute a “test” for depression (often requested by families), but they continue to be of research interest to better understand the disorder and obtain potential markers for treatment response.

Panic disorder has been linked to an exaggerated fear of suffocation due to a biologic and possibly heritable hypersensitivity in the brain’s receptors for carbon dioxide [reviewed in (21)]. Differential sensitivity in the receptors of anxious versus control children is shown by carbon dioxide panic provocation challenge studies; evidence of higher than average incidences of anxiety disorders in asthmatic children; and lower incidences of anxiety symptoms in children with faulty carbon dioxide receptors. It is possible that these overly sensitive receptors trigger isolated panic attacks, which are thought to cue conditioning experiences and in turn phobic avoidance.

“Anxiety sensitivity” has been found to be a specific predictor of onset of adolescent panic [reviewed in (22)]. This construct, measurable by self-report inventories, consists of the tendency to respond with anxiety to sensations of autonomic arousal. To address this response in the cognitive-behavioral treatment of panic, patients are helped to reinterpret sensations of autonomic arousal as benign rather than threatening.

Several medical conditions have been linked to childhood anxiety or depression [see (23)]. The sudden onset of anxiety or depression in a previously outgoing child should alert the clinician to explore a medical or traumatic etiology. Hyperthyroidism commonly produces anxiety states, and hypothyroidism can contribute to depression. Medications that are sympathomimetic (for example, Ventolin inhalers for controlling asthma) can contribute to anxiety. Endocrine conditions affecting adrenal steroids have been linked to mood disorders. In girls, consumption of caffeinated beverages, particularly diet colas, often contributes to anxiety. Many over-the-counter diet aids also contain stimulants that can exacerbate anxiety. Pediatric
chronic illnesses have also been linked to depression, with both child and illness characteristics contributing to the vulnerability [reviewed in (24)].

Turner syndrome is a condition found in phenotypic girls who have monosomy of the X chromosome. It is characterized by gonadal failure, short stature, and some associated medical abnormalities. Many girls with Turner’s syndrome suffer symptoms of anxiety and depression [reviewed in (25)]. In the past, these symptoms were thought to reflect low self-esteem secondary to the condition. Recent studies, however, suggest that these symptoms persist even when affected girls are treated with growth hormone to address short stature. Differences in executive cognitive functions and brain anatomy distinguished monozygotic twins discordant for Turner syndrome (26), and may account for difficulty modulating negative affect in this population.

Attachment theory proposes that infants are predisposed to behave in ways that enhance proximity to their caregivers, and caregivers are prone to behave reciprocally (27). When parent-infant attachment is secure (as demonstrated in the laboratory), the primary caregiver eventually serves as a “secure base” from which the infant can explore the world, returning at times of distress. Insecure attachment, however, has been linked with childhood anxiety symptoms concurrently and with anxiety disorders prospectively (28). Security or insecurity of attachment has been found to be relatively independent of child temperament. Interventions that promote secure attachment relationships therefore have the potential to prevent some anxiety problems in temperamentally vulnerable children.

The influence of early attachment on brain development has been the subject of increased study recently [reviewed in (29)]. Studies of animals and human beings have both confirmed that adverse early relational experiences can result in activation and eventually alteration of the functioning of the hypothalamic-pituitary-adrenal (HPA) axis. Physiologic responses to stress are thus changed. Furthermore, the altered HPA functioning influences the development of brain pathways linked to anxiety and depression, increasing vulnerability to these disorders. Maternal depression is known to be associated with high rates of insecure attachment and risk of adolescent depression. While modeling of depressive cognitions and impaired parenting likely contribute to this association, the findings suggest that alterations of the HPA axis may also play a role.

Depressed adults regularly report more conflict than do normal adults as well as more rejection, more communication problems, and less support in their childhood families of origin [reviewed in (12)]. Given that these reports are retrospective, however, it is unclear to what extent they are biased by the cognitive distortions associated with depression. In depressed children, the possibility of bidirectional effects must also be considered. For example, a sullen, irritable child with depression may elicit rejecting or conflicted interactions with parents, but these interactions may, in turn, worsen the child depression. Children of depressed
mothers may be at particular risk, however, both because of modeling of depressive behaviors and cognitions and because of high rates of insecure attachment (see above).

In children with anxiety disorders, observational studies of parent-child interactions have supported a link between anxiety and overprotective parenting. For example, Hudson and Rapee (30) found that mothers of anxiety-disordered children were more controlling and intrusive during a puzzle task than mothers of nonanxious children. They proposed a bidirectional relationship such that the parent of an anxious child may be more likely to become overinvolved with the child in an effort to reduce and prevent the child’s distress. In turn, this behavior would reinforce the child’s anxiety by promoting beliefs that the world is a dangerous place from which the child needs protection and over which he or she has no control. The parent’s own anxiety is also likely to compound this maladaptive pattern of behavior.

Parental modeling has also been implicated in the development of anxiety disorders in children. For example, adult social phobics have retrospectively reported their families as encouraging isolation and constraining contacts with neighbors, relatives, and acquaintances [see (31)]. In addition to minimizing or discouraging social networking, parents may also provide children with information that heightens anxiety about certain situations.