Developmental Psychopathology

Suniya S. Luthar

Rebecca P. Prince

Developmental Psychopathology Defined: Major Features

Developmental psychopathology is an integrative discipline, wherein principles from classical developmental theory are applied to investigate clinical and psychiatric phenomena (1,2,3,32). This integration of perspectives is invaluable because it promotes our understanding of atypical development and also illuminates understanding of normative developmental processes. To illustrate, applications of developmental theories such as those of Werner, Piaget, and Erikson provide critical insights into the organization and causes of different forms of maladjustment. Conversely, studies of pathology enhance our knowledge of normal development, particularly in terms of individual differences in development as well as risk and protective processes associated with different types of outcomes.

Whereas developmental and clinical psychology are integral elements in the field of developmental psychopathology, the scope of this integrative discipline extends beyond these areas. Theory and methods from these domains are integrated with those from various others, including epidemiology, biology, neuroscience, sociology, and anthropology. Such multidomain, multicontextual approaches to inquiry are essential in moving toward the long-term goal of a more comprehensive understanding of the development of psychopathology.

A final feature of developmental psychopathology is that it bridges the often wide span between empirical research and the application of knowledge, to benefit at-risk populations. Investigators in this tradition design and implement interventions that are based in developmental theory and research on risk and protective processes, such that they inform both preventive interventions and social policy.

To summarize, the four central characteristics that define the field of developmental psychopathology are 1) the use of classical developmental theory and research to inform issues of psychopathology, 2) the use of insights from at-risk or atypical populations to increase our understanding of normal developmental processes, 3) integration of developmental and clinical perspectives with those from other disciplines, and 4) the derivation of implications for preventive and therapeutic interventions, and for social policy.

Risk

In developmental psychopathology research, risk is defined in terms of statistical probabilities: A high-risk condition is one
that carries high odds for measured maladjustment in critical domains (5). Exposure to community violence, for example, constitutes high risk given that children experiencing it reflect significantly greater maladjustment than those who do not (6). Similarly, maternal depression is a risk factor in that children of mothers with depressive diagnoses can be as much as eight times as likely as others to develop depressive disorders themselves by the adolescent years (7).

In addition to establishing discrete risk dimensions such as community violence, poverty, or parent mental illness, researchers have also examined composites of multiple risk indices such as parents’ low income and education, their histories of mental illness, and disorganization in their neighborhoods. Seminal research by Rutter (8) demonstrated that when risks such as these coexist (as they often do, in the real world), effects tend to be synergistic, with children’s outcomes being far poorer than when any of these risks existed in isolation. Use of this cumulative risk approach is well exemplified in work by Sameroff and his colleagues (9,10). These authors computed a total risk score across 10 different dimensions, assigning for each one, a score of 1 (versus 0) if the child fell in the highest quartile of continuous risk dimensions, and for dichotomous dimensions such as single parent family status, if they were present in that child’s life. An alternative approach, exemplified in work by Masten and her colleagues (11), involves standardizing values on different risk scales and adding them to obtain a composite.

Decisions regarding the use of single- or multiple-risk indices in resilience research depend on the substantive research questions. The former is used, obviously, when applied researchers seek to identify factors that might modify the effects of particular environmental risks known to have strong adverse effects, so as to eventually derive specific directions for interventions. Examples are parental divorce or bereavement; knowledge of what ameliorates the ill effects of these particular adversities has been valuable in designing appropriate interventions (12,13). Additive approaches are more constrained in this respect, precluding identification, for example, of which of the indices subsumed in the composite are more influential than others. On the other hand, composite risk indices generally explain more variance in adjustment than do any of them considered alone, and as noted earlier, they may be more realistic in that many of these risks do cooccur in actuality (5,14).

Risk is rarely absolute; the potential for deleterious outcomes varies according to age as well as other child characteristics. Prolonged separation from the primary caregiver, for example, is more harmful for infants and toddlers than for older children, whereas community violence is less likely to affect preschoolers than older youth who are more able to move about the neighborhood independently. By the same token, there are some risks relatively unique to particular groups. An example is racial discrimination, which affects ethnic minority groups but not children of Caucasian heritage.

The same construct can connote risk in one setting but be relatively benign or even beneficial in others. An example is stringency of parent discipline. Whereas high levels of control and strictness are often seen as deleterious for children, a series of studies have shown that they are actually beneficial for youngsters living in dangerous inner city neighborhoods (15,16,17,32).

Disorder

In developmental psychopathology as in child psychiatry, the notion of disorder often represents psychiatric diagnoses. Researchers typically assess diagnoses via structured interviews such as the Schedule for Affective Disorders and Schizophrenia for School-Aged Children (K-SADS-PL) (19) or the Diagnostic Interview Schedule for Children (NIMH DISC-IV) (20), which are usually administered to the child aged 5 and older as well as the primary caregiver. For each diagnostic category, these interviews have a series of initial probes to determine the existence of a disorder, and if responses are in the affirmative, then additional probes are asked to determine if diagnostic criteria are met.

The other approach, also commonly used, is to assess overall children’s symptom levels on different maladjustment domains, via instruments such as the Behavior Assessment System for Children, (BASC) (21) or the Child Behavior Checklist (CBCL) (22) [and its variants, the Teacher Rating Form (TRF) and the Youth Self Report (YSR) (23)]. These measures include a list of symptoms from diverse maladjustment domains which collectively yield scores on discrete subscales (such as attention, conduct, or depressive problems); in turn, composite scores across related subscales indicate overall maladjustment, such as internalizing and externalizing symptoms (CBCL), or overall dimensions of behavior and personality (BASC).

Such dimensional measures have two major advantages; they capture a wide range of functioning and are very well normed. With regard to the first of these features, symptom scales such as the CBCL and BASC characterize children in terms of varying severity of dysfunction as opposed to simply the presence or absence of diagnoses. From a research standpoint, this is a major advantage because the greater the variance on a particular dimension, the more likely it is that it will show statistical links with other constructs (such as potential causes or ramifications of the symptoms). The issue of norms, similarly, is critical in gauging children’s adjustment levels relative to those of the average child of the same age. Instruments such as the CBCL have been administered to thousands of children from all over the country (and world) and as a result, we know the average symptom levels on these. Typically, average levels of problems correspond to a T score of 50 with a standard deviation of 10. Thus, if a child were to obtain a score of over 65 on the YSR, this would represent “much above average” dysfunction and a T score of 70 or more would indicate problems “very much above average.”

Making such judgments about functioning vis-ā-vis the average child is much more complicated with psychiatric disorders. There have been several large-scale epidemiological studies on children’s diagnoses, but there is some disagreement on rates of different disorders, with variations, for example, with the particular structured interview used. To illustrate, an NIMH study using the DISC reported that about 6% of adolescents suffer from depression (24), whereas a study using the K-SADS found the point prevalence to be 2.9% (25). At the same time, structured interviews remain the method of choice when the goal is specifically to assess the incidence of actual psychiatric diagnoses, rather than severity of symptoms.

A thorny problem in assessing childhood disorders— regardless of whether the approach involves diagnoses or symptom levels—is that there is considerable disagreement among respondents. The kappa statistic is commonly used to assess agreement across raters, with values above .75 representing high levels of agreement, values in the range of .40 to .75 representing moderate levels of agreement, and values below .40 representing low levels of agreement (26). In studies involving psychiatric diagnoses, agreement rates between parents and children have ranged from k = .32 for diagnoses of separation anxiety to k = .22 on diagnoses of general anxiety disorder to as low as k = .17 on ADHD (27,28). Further demonstrating the disconnect between the child’s and parents’ understanding of the child’s functioning, Roberts and colleagues (29) found that in measuring overall mental health, life satisfaction, happiness, and role strain, parent–child agreement was never above .20. On dimensional measures, similarly, the
correlation between parents’ reports and children’s reports has ranged from k = .02 for anxious/depressed symptoms to k = .14 for delinquent behavior (30); there is generally more agreement among mother and father, but again, low consensus between parents and teachers (31).

Researchers have dealt with this disagreement in various ways. In the case of psychiatric diagnoses, a common approach is to use the “either/or” rule, assuming the child does in fact have a diagnosis if either the parent or the child indicates this is the case (32). An alternative strategy is by prioritizing adults’ reports for some domains (conduct problems) and children’s for others (depression), with the rationale that children are likely to underreport their own oppositional behavior, for example, or that parents are likely to know less about their child’s inner life (33,34). Still others have separately considered parents’ and children’s reports, with the rationale that each reveals information not captured by the other (35,36,37).

A final comment about measurement of disorder: What is considered abnormal in one setting may be normative or even adaptive in another. This point is well illustrated in a paper by Richters and Cicchetti (38) entitled Mark Twain meets DSM-III-R: Conduct disorder, development, and the concept of harmful dysfunction, with regard to definitions of conduct disorders. The authors argue that many inner city youth might meet DSM criteria for conduct disorders but in their own subculture, being able to defend themselves physically can be quite adaptive. Accordingly, they exhort additional consideration of the notion of “harmful dysfunction,” put forth by Wakefield (39,40,41). Wakefield argued that the DSM definition of mental disorders fails to distinguish adequately an individual’s negative reactions to his problematic environment from a “true” mental disorder, and that a mental disorder is best conceptualized as a harmful dysfunction, where harm is a value judgment regarding the undesirability of a condition, and dysfunction is the failure of a system to function as designed by natural selection. Anchored in notions of evolutionary design, Wakefield specifically defines a harmful dysfunction as the “harmful failure of an internal mechanism to perform a natural function for which it was biologically designed” (42).*

Resilience

A salient construct in developmental psychopathology is resilience: a phenomenon or process reflecting relatively positive adaptation despite experiences of significant adversity or trauma. Inherent in this definition lie two fundamental conditions: significant risk (or adversity) and positive adaptation. Thus, resilience is never directly measured, but is indirectly inferred based on evidence of the two subsumed constructs.

The notion of risk has already been discussed; positive adaptation, the second element in the construct of resilience, is defined as an outcome that is substantially better than what would be expected with respect to the risk circumstance being studied. In past studies of resilience across diverse risk circumstances, positive adaptation has been defined in terms of behaviorally manifested social competence, or success at meeting stage-salient developmental tasks (5,14,43). Among young children, for example, competence was operationally defined in terms of the development of a secure attachment with primary caregivers (44), and among older children, in terms of aspects of school-based functioning such as good academic performance and positive relationships with classmates and teachers (43,45).

In addition to being developmentally appropriate, indicators used to define positive adaptation must also be conceptually of high relevance to the risk examined in terms of both domains assessed and stringency of criteria used (46). When communities carry many risks for antisocial problems, for example, the degree to which children are able to maintain socially conforming behaviors is an appropriate indicator of success (47), whereas among children of depressed parents, the absence of depressive diagnoses would be of special significance (48,49). With regard to stringency of criteria, similarly, decisions must depend on the seriousness of the risks under consideration. In studying children facing major traumas, it is entirely appropriate to define risk-evasion simply in terms of the absence of serious psychopathology (psychiatric diagnoses) rather than superiority or excellence in everyday adaptation (50).

Regardless of whether competence is described as risk evasion or positive adaptation, competence must be defined across multiple spheres. The multilevel measurement of competence, therefore, differs from the measurement of risk, which may legitimately involve one or multiple negative circumstances. However, doing well in only one domain cannot be conceptualized as connoting resilience, as overly narrow definitions can fallaciously convey a picture of “success in the face of adversity.” Adolescents, for example, might be viewed very positively by their peers but at the same time, perform poorly academically, or even demonstrate conduct disturbances (51,52), such that peer popularity by itself cannot be seen as an indicator of overall risk evasion.

The major focus of resilience researchers is to identify vulnerability and protective factors that might modify the negative effects of adverse life circumstances, and having accomplished this, to identify mechanisms or processes that might underlie associations found (5,53,54,32). While researchers have debated how to delineate such factors statistically (see (14,56)), the conceptual definitions are fairly straightforward. Vulnerability factors or markers encompass those indices that exacerbate the negative effects of the risk condition while protective factors modify the effects of risk in a positive direction.

In some instances, it is more appropriate to define positive adaptation in terms of the family or community rather than necessarily of the child him- or herself. As Seifer (57) has argued, because infants’ and toddlers’ functioning is often regulated by their caregivers, it can be more logical to operationalize positive adjustment in terms of the mother–child dyad or family unit rather than in terms of the young child’s behavior. Similarly, there are times when the label resilience is most appropriate for communities of well functioning at-risk youth. Research on neighborhoods, for example, has demonstrated that some low-income urban neighborhoods reflect far higher levels of cohesiveness, organization, and social efficacy than others (58,59), with the potential, therefore, to serve as important buffers against negative socializing influences.

Factors Affecting Resilience and Vulnerability

In discussions that follow, we overview major factors that contribute to relatively positive or negative outcomes among at-risk children. These factors fall into three broad categories: aspects of families, features of communities, and attributes of the children themselves.

Familial Factors in Resilience and Vulnerability

The critical importance of strong family relationships has been emphasized in various studies of resilience; a plethora of studies identify supportive and responsive parenting as being among the most robust predictors of resilient adaptation (5,8,54,56,60,61,62,63,64). In particular, early family relationships are extremely important in shaping long-term resilient trajectories. In their comprehensive review of the early childhood literature, Shonkoff and Phillips (65) emphasized that “[relationships] that are created in the earliest years … constitute a basic structure within which all meaningful development unfolds.” Early experience places people on probabilistic trajectories of relatively good or poor adaptation, shaping the lens through which subsequent relationships are viewed and the capacity to utilize support resources in the environment. Thus, if early attachments are insecure in nature, at-risk children tend to anticipate negative reactions from others and can eventually come to elicit these; these experiences of rejection further increase feelings of insecurity (66,67,68). Conversely, at-risk children with at least one good relationship are able to take more from nurturing others subsequently encountered in development (44,65,67,32). The protective potential of strong family relationship has been demonstrated for mothers, fathers, and siblings (70,71).

Only gold members can continue reading. Log In or Register to continue