The following reports are based on a symposium held in New York City
at the annual meeting of the American Academy of Child and Adolescent Psychiatry.

A half century after the classic behavioral studies of Bowlby and Spitz, the field of early childhood psychiatry is still, in many respects, in its infancy. Numerous basic issues remain unresolved. Some experts, for example, question whether behavioral symptoms in this population represent actual psychiatric disorders or should be viewed merely as markers for possible future psychopathology. And even if a clinician accepts in principle that these symptoms constitute true disorders, diagnosis is hindered by the fact that infant-specific criteria are still being conceptualized. Needless to say, reliable treatment data are essentially nonexistent.

At the annual meeting of the American Academy of Child and Adolescent Psychiatry, researchers described some of the challenges in identifying psychiatric disorders in infants and preschoolers and reviewed initial efforts to define and validate appropriate criteria.

DIAGNOSING DEPRESSION IN PRESCHOOLERS

The idea that very young children are capable of experiencing depressed affect is “intuitively difficult for us to grasp,” acknowledged Joan L. Luby, MD, Assistant Professor of Psychiatry at Washington University School of Medicine, in St. Louis. However, evidence indicates that preschoolers and perhaps even infants do experience depressive disorders, and it may be especially important to diagnose and treat such cases in order to minimize the likelihood that the disorder will become chronic and relapsing, Dr. Luby emphasized. “Neuroplasticity of the brain is much greater in the first four to five years of life, suggesting that behavioral and [especially] pharmacological interventions could be more effective and actually brain changing when applied at this time,” she noted.

However, identifying major depression and related disorders in preschoolers requires the development of age-appropriate diagnostic criteria. Empirical studies in the 1970s and 1980s that used DSM criteria identified depression in so few children under the age of 6 that the researchers concluded depression may not exist in the very young. “I think that the error in those studies is the use of the DSM criteria,” Dr. Luby said. “These criteria often just do not apply to the life experiences of preschoolers. For example, preschoolers don’t have impairment in the workplace.”

In fact, she estimated that the DSM-IV criteria do not identify 85% of preschool children who meet broader criteria for depressive disorders (described below) and who have multiple clinical markers. A further problem, according to Dr. Luby, is that in some cases strict application of current criteria may result in misdiagnosis. “I do think we’re diagnosing attention-deficit/hyperactivity disorder (ADHD) in some kids who could better be understood as having anxiety and depressive disorders,” she said.

AN ALTERNATIVE APPROACH TO DIAGNOSIS

In an ongoing study that has thus far involved about 100 children, Dr. Luby and her colleagues have been working to devise developmentally appropriate diagnostic criteria for major depression in preschoolers. In particular, the researchers have been investigating a modified version of the DSM-IV criteria in which children need only have four symptoms instead of five. Equally important, the new criteria eliminate the requirement that the patient have had two weeks of stable depressed mood, the reasons being, according to Dr. Luby, that it is very unusual to find a preschooler with “a stable anything” over a two-week period, let alone a stable mood, and that adults tend not to be sufficiently aware of children’s internal experiences to judge their mood at this level of detail.

Do these modified criteria identify a clinically valid population? The researchers are encouraged by preliminary data showing that preschoolers “who meet these modified criteria had significantly greater family histories of affective disorders—major depression, bipolar disorder, and anxiety disorders—compared to normal controls,” Dr. Luby reported. The depressed children were also more likely than children with ADHD to have a family history of affective disorders, although familial rates of major depression per se did not differ between the two groups.

In addition, the researchers are investigating potential biologic and neuropsychiatric markers for depression. Although it is too early to draw any firm conclusions, Dr. Luby noted that preliminary measurements of salivary cortisol indicate that the depressed children, normal controls, and children with ADHD all exhibit different patterns of cortisol release when exposed to stressful circumstances—such as separation from their primary caregiver—during an observational session. “The depressed group appears to come in high, go down very low, and then go up again,” whereas children with ADHD show an almost inverse pattern, Dr. Luby reported. This hyperreactivity to stress in depressed children mirrors the findings seen in depressed adults assessed with the dexamethasone suppression test.

Another possible marker may be impaired visuospatial skills. Preschoolers with depression, like their adult counterparts, appear to perform more poorly than controls on visuospatial tasks. However, they recognize the emotional intensity of sad female faces more accurately than do healthy children, a finding that contrasts with analogous studies in adults.

The researchers have also found that, according to data gathered using the Child Behavior Checklist, depressed preschoolers have far more internalizing symptoms than both controls and children with ADHD, and more externalizing symptoms than the control group. “You wouldn’t expect them to have more externalizing symptoms than the externalizing (ie, ADHD) group,” Dr. Luby noted. Furthermore, findings from a puppet-based interview indicate that depressed children report lower levels of happiness than do controls. (Children of this age do not acknowledge feeling sad but do describe themselves as feeling “less happy,” according to Dr. Luby.)

In general, then, “there does appear to be preliminary validation for early forms of [depressive] disorders,” Dr. Luby concluded. Moreover, “the fact of greater brain neuroplasticity in children under five is a promising suggestion that early intervention is important.” Unfortunately, thus far “there is absolutely no data” regarding the efficacy of pharmacologic or behavioral treatments in this population, although Dr. Luby believes both approaches will prove useful.

Suggested Reading
Birmaher B, Ryan ND, Williamson DE, et al. Childhood and adolescent depression: a review of the past 10 years. Part I. J Am Acad Child Adolesc Psychiatry. 1996;35:1427-1439.

DIAGNOSING PTSD IN INFANTS

The shortcomings of DSM-IV in diagnosing early childhood psychopathology also apply to posttraumatic stress disorder (PTSD), according to Michael S. Scheeringa, MD, Assistant Professor of Psychiatry and Neurology and Clinical Assistant Professor of Pediatrics at Tulane University School of Medicine, New Orleans. New criteria devised by Dr. Scheeringa and his colleagues appear to be more developmentally sensitive than those specified by DSM-IV, although preliminary data indicate that clinicians must still rely on parental reports to determine the presence of most symptoms.

About half of the new criteria, which are described in the January Journal of the American Academy of Child and Adolescent Psychiatry, are essentially identical to criteria specified by DSM-IV. However, several others have been substantially modified from their DSM-IV analogs (eg, according to the new guidelines, recurrent recollections need not be distressing to the child) and a few items are entirely new (eg, the development of new fears). The researchers have been testing the procedural, criterion, and discriminant validity of the new criteria in an ongoing study of children under age 4 who were severely traumatized by sexual or physical abuse, witnessing domestic violence, undergoing cancer treatment, dog attacks, or other causes. Under the study protocol, the assessment of each child takes about two hours and includes a 60-minute interview with a parent (while the child plays in the room), two brief play sessions (child–parent and child–examiner), and a discussion with the parent regarding his or her own PTSD symptoms. “These are pretty practical sequences,” noted Dr. Scheeringa. “These are things that you would do in your own office.”

The other part of the evaluation is a 10-minute session in which the trauma is reenacted under the examiner’s guidance using props and dolls. If the child had been in a car accident, for example, the clinician might reenact the accident using a toy car or prompt the child to recreate the accident himself.

Two clinicians viewed videotapes of each evaluation to determine which of the modified and DSM-IV criteria were present; a third researcher settled any differences of opinion that arose. As part of their effort to assess the procedural validity of the criteria, the researchers tracked how many PTSD symptoms could be observed firsthand during evaluations. Unfortunately, the answer seems to be “very few,” according to data from the first 15 traumatized children (and from 12 comparison subjects recruited from an inner-city pediatric clinic). Of the 26 symptoms specified by either DSM-IV or the alternative criteria, only eight were observed firsthand in one or more patients; of the 132 PTSD symptoms identified in the entire group of traumatized children, only 16 (12%) were apparent during visits. Thus, clinicians had to rely on parental report for the other 88% of symptoms, Dr. Scheeringa noted.

ARE THE NEW CRITERIA SUPERIOR?

Nine of the 15 traumatized children met the modified PTSD criteria, whereas only three met the DSM-IV criteria; six others fell slightly short of satisfying the DSM-IV specifications, typically because a single symptom or cluster did not apply. Which approach had superior criterion validity? Several clues point to the new criteria, including the fact that the nine patients with the greatest number of symptoms all qualified for a PTSD diagnosis under the new guidelines. “That was not the case with DSM-IV,” Dr. Scheeringa noted; three of the five children who had the most DSM-IV symptoms did not receive a diagnosis of PTSD.

Both sets of criteria proved to have good discriminant validity: the average number of symptoms identified in traumatized children (6.2 for DSM-IV, 7.5 for the new criteria) was far higher than the corresponding values for the control sample (1.2 and 1.5, respectively). This might seem like an obvious finding, Dr. Scheeringa noted, but it is an important one from a diagnostic standpoint because about two thirds of PTSD symptoms (eg, sleep problems) are nonspecific.

Even if the new criteria do represent an improvement, establishing the presence of a particular symptom remains a challenge. Indeed, the researchers identified three common obstacles to establishing that a PTSD symptom was present:

• symptoms were often difficult to differentiate from age-appropriate behaviors (eg, temper tantrums);
• the threshold for impairment was sometimes uncertain (eg, how many tantrums per week are indicative of a problem?); and
• in some instances, the timing of symptom onset was not clear, so that the researchers could not reliably determine whether the symptom’s appearance preceded or followed the traumatic event.

MORE TO COME

The investigators are planning to increase the study population to 60 traumatized children, each of whom will be followed for two years. The expanded study will also examine several additional variables, including parent–child interactions, parental symptoms, and autonomic reactivity. The latter is of interest because of the oft-reported finding that the heart rates of adults with PTSD quicken when they think about a traumatic experience; it is not known, however, if the same is true of children. In the Tulane study, preschoolers sit in a chair for 20 minutes, hooked up to ECG electrodes, while the researchers ask them to describe a remote memory (such as visiting the zoo), to talk about and/or reenact their traumatic experience, and to listen to their mother describe the event. Preliminary findings suggest that the increase in autonomic activity may be less robust in children than it is in adults. Dr. Scheeringa quipped, however, that he doesn’t care what the findings show, given the difficulty of performing ECGs with preschoolers. “I’m just happy they don’t tear the electrodes off.”

Suggested Reading
Scheeringa MS, Peebles CD, Cook CA, Zeanah CH. Toward establishing procedural, criterion, and discriminant validity for PTSD in early childhood. J Am Acad Child Adolesc Psychiatry. 2001;40:52-60.

ATTACHMENT DISORDERS IN INFANCY AND EARLY CHILDHOOD

A recent study suggests that attachment disorders can be reliably diagnosed in very young children, although at least one such disorder proved difficult to assess in this population, reported Neil W. Boris, MD, an Assistant Professor in the Departments of Psychiatry and Community Health Sciences at Tulane. Not surprisingly, the findings also indicate that children who are placed in foster care after being abused and neglected are at extremely high risk of developing an attachment disorder.

The phenomenon of attachment is operative throughout a person’s life but typically manifests around the sixth or seventh month after birth; in infants, it can be conceptualized as the balance between the urge to explore and the desire to remain near the caregiver (which provides a sense of security). Although attachment theory has advanced substantially over the past half century, progress has been much slower on the clinical front: There is still no consensus regarding “what defines an attachment disorder [and] what requires treatment from clinicians,” Dr. Boris noted. At present, most experts view the attachment disorder spectrum as including disrupted attachment disorder, which often occurs in children who are removed from their primary caregiver; secure base distortions, such as role reversal; and disorders of nonattachment, which can be divided into inhibited and disinhibited subtypes.

To determine whether attachment disorders can be diagnosed reliably in high-risk children under age 4, Dr. Boris and colleagues recruited a study cohort designed to mimic real world clinical populations: The children either had been placed with foster parents due to serious abuse and neglect or they lived in a homeless shelter and had relatively young parents (ages 18 to 21) who were raised in housing developments or other high-risk environs. “These are the crème de la crème of Louisiana’s maltreated children under 4,” Dr. Boris said. A third group of children, who participated in a Head Start program for children of low socioeconomic status, served as a comparison group.

As part of the evaluation, each child participated in a 25-minute observational assessment; in addition, the researchers conducted a semistructured interview with the child’s caregivers. A panel of three clinicians trained in the diagnosis of attachment disorders reviewed tapes of the observational session and interview and determined whether an attachment disorder should be diagnosed.

DO THE RATERS AGREE?

In cases where an attachment disorder was not diagnosed, interrater agreement was fairly substantial, resulting in a k value of 0.69, Dr. Boris reported. “Given that we’re early in the process, we felt that this was a fairly reasonable mark to have hit,” he said. Interrater reliability was also quite high for diagnosing disrupted attachment disorder, but it was less so for self-endangering secure base disorders. In general, agreement between raters was moderate for reactive attachment disorder and nonattachment disorders, although it sometimes varied substantially depending on the diagnostic criteria used; agreement was rather poor, for example, when the raters used the International Classification of Disease criteria for the indiscriminate subtype of nonattachment disorder. There were too few cases of inhibited nonattachment disorder to determine interrater reliability.

At least one attachment disorder diagnosis, role reversal, proved to have little clinical utility in this population: A remarkable 25% of the total sample met current criteria. Not unexpectedly, “all three clinicians involved commented that the threshold for this disorder was too low,” Dr. Boris reported. These diagnoses often stemmed from caregiver reports that their children, even before the age of 18 months, offered comfort or brought tissues when their parents were upset. “One would suggest that that’s somewhat precocious given what we know about empathy and infancy,” noted Dr. Boris. Because the Strange Situation procedure classified most of these children as securely attached, the researchers decided that role reversal lacked validity under current criteria and excluded the diagnosis from the study’s other analyses.

Further examination of the Strange Situation data revealed that 85% of the subjects who were securely attached did not have an attachment disorder. However, two securely attached children met criteria for the indiscriminately social form of nonattachment disorder; this finding did not surprise Dr. Boris, because the children’s history suggested that they “would go to the mailman if we were to bring him into the laboratory.” Such cases indicate that “there is some complexity with regard to how one looks at the [Strange Situation] data in these classification procedures.”

Finally, the study provided confirmation that early childhood experiences play a key role in attachment. The researchers identified attachment disorders in nearly two thirds of children in foster care and 20% of those who lived in a homeless shelter, but none of the children from the Head Start program. “Social risk and previous experiences, as we have always thought, are indeed related to attachment disorders,” Dr. Boris noted.

Summing up the findings, Dr. Boris concluded that “attachment disorders can be reliably diagnosed [in young children], but some appear to be more clinician-friendly than others.” Some criteria need to be refined, he said, and the diagnosis of role reversal should be reconceptualized. “As a clinician, [I believe that] this disorder likely exists, but perhaps more in children past 4 years of age and in certain family situations.”

—Kathryn Blair

Suggested Reading
1.Boris NW, Wheeler EE, Heller SS, Zeanah CH. Attachment and developmental psychopathology. Psychiatry. 2000;63:75-84.
2. Boris NW, Zeanah CH, Larrieu JA, et al. Attachment disorders in infancy and early childhood: a preliminary investigation of diagnostic criteria. Am J Psychiatry. 1998:155:295-297.

Return to table of contents