ارزیابی مدل های رابطه آسیب شناسی روانی شخصیت در کودکان و نوجوانان

Connections between personality traits and psychopathology in children and adolescents have frequently been reported in research studies. However, despite the occurrence of significant and systematic relationships between personality and mental disorders in childhood, a thorough understanding of the cause, nature, and implications of these relationships is lacking. In this paper, a comprehensive taxonomy of childhood personality is used to link research on children with that on adults, as well as provide a framework for discussing the personality–psychopathology relationship. Next, research on children and adolescents is integrated into various proposed models of the personality–psychopathology relationship. Finally, clinical implications and future directions are proposed for research on personality and psychopathology in children. Hypotheses regarding connections between personality and psychopathology have a long-standing history, although ideas about the nature of these connections have changed over time (Maher & Maher, 1994). At the root of most of these hypotheses is the idea that psychopathology occurs in individuals within the context of premorbid personality, and understanding the connections between personality and psychopathology can lead to increased understanding of the individual's functioning. Over the last two decades, a growing body of research has attempted to explain the nature of these relationships (see Krueger and Tackett, 2003 and Widiger et al., 1999 for reviews), building on current research investigating the etiology and structure of both personality and psychopathology. However, this work has primarily focused on adult populations, despite the growing evidence for robust associations between personality traits and mental disorders in children and adolescents. In order to promote understanding and further investigation of the relationship between personality and psychopathology, existing work with children and adolescents must be integrated into research on adults to create a broader developmental picture. In addition, use of a common language regarding personality structure and models of personality–psychopathology relationships across various ages will facilitate ultimate understanding of these relationships across the lifespan.

The goal of the present paper is to review evidence for various proposed models of the relationship between personality and psychopathology in children and adolescents. Before specifically discussing the personality–psychopathology relationship, it is necessary to clarify the personality constructs discussed here. In particular, some continuity in discussion of personality and temperament constructs and structure across ages is necessary to understand and integrate research extending from childhood to adulthood. Questions of personality stability are also discussed as they relate to this issue. Thus, I begin by defining personality and the framework that will be used to organize the personality–psychopathology literature reviewed here. 1. Defining personality in childhood and adolescence 1.1. Temperament versus personality In adults, individual differences in one's characteristic ways of behaving, thinking, and feeling are typically defined as personality. However, in children, characteristic individual differences may be described as temperament traits as well as personality traits. The study of temperament typically refers to traits or characteristics that are biological in nature and appear very early in life ( Frick, 2004, Goldsmith et al., 1987 and Rothbart and Bates, 1998). Temperament is often described as a subset of personality, with personality referring to a broader realm of individual characteristics ( Rothbart et al., 2000, Rothbart and Bates, 1998 and Shiner and Caspi, 2003). Furthermore, temperament is thought to be most directly observable during infancy and toddlerhood ( Goldsmith et al., 1987) and to make up the entirety of personality during these early years ( Shiner & Caspi, 2003). As children develop, it is likely that early temperamental traits develop into broader, more inclusive higher-order personality traits as well as increasingly differentiated lower-order traits (Buss and Finn, 1987, Caspi et al., 2005, Lahey, 2004, Rothbart and Ahadi, 1994, Shiner, 1998 and Shiner, 2006). Specifically, personality develops as children progress through various cognitive and emotional stages that increasingly allow them to interact with, experience, and respond to the world in more complex ways (Caspi, 2000 and Rothbart and Ahadi, 1994). Thus, the structure of personality may change as children gain new skills (e.g., motor or language), the capacity to regulate emotions (e.g., ability to inhibit antisocial responses), and a sense of self (Rothbart and Ahadi, 1994, Shiner, 1998 and Shiner et al., 2002). However, despite growing consensus that temperamental traits make up the core of later personality, a clear understanding of the developmental relationship between temperament and personality is lacking (Halverson et al., 2003 and Shiner and Caspi, 2003). One recent study investigated the relationship of an early temperamental trait (inhibition to novelty) to a later personality trait (inhibitory control) over infancy to early childhood, which is considered a key developmental period for later personality traits (Aksan & Kochanska, 2004). Similar work is needed regarding other temperament and personality traits in order to understand the nature of development for important individual differences. 1.2. Measurement of temperament and personality The work of Thomas and Chess (1977) is considered the pioneering effort in measuring temperamental characteristics in infancy and childhood. Thomas and Chess defined a model of temperament based on an inductive content analysis of parent interviews about their infant's behavior. The resulting model was made up of nine dimensions that had been identified as behaviors with potential significance for psychological development. While this work was greatly influential in current conceptualizations of temperament measurement, psychometric limitations of the model have since been identified (e.g., Halverson et al., 2003). These limitations (such as a lack of discriminant validity among the dimensions) and debate over the appropriate number of dimensions led researchers to develop alternative models (Rothbart, 2004 and Rothbart and Bates, 1998). Presently, numerous frameworks exist for the structure of temperament and childhood personality, with researchers using various measures of temperament constructs. While research on temperament has increased since the work of Thomas and Chess, there remains a lack of consensus regarding the best framework (Goldsmith et al., 1987 and Rothbart and Bates, 1998), which has resulted in a somewhat fragmented literature. However, the apparent fragmentation of temperament research may to some extent reflect a tendency for researchers to give similar constructs different names between models more than it reflects a lack of agreement over the primary constructs to study (Rothbart et al., 2000). Nevertheless, the field of temperament and childhood personality needs a unifying framework to allow organization and integration of empirical findings and facilitate communication among researchers within the field, as well as with researchers in other disciplines (Halverson et al., 2003, Lonigan et al., 2004 and Shiner and Caspi, 2003). The adult personality literature has largely converged on the Five Factor Model (FFM) as an integrating framework. The factors in the FFM are: Extraversion, Agreeableness, Neuroticism, Conscientiousness, and Openness to Experience. Extraversion is a trait indexed by characteristics such as sociability, positive energy, and gregariousness. Agreeableness is represented by characteristics such as empathy and warmth toward others. Neuroticism is related to an individual's tendency to experience feelings of anxiety, irritability, and depression. Conscientiousness reflects characteristics of organization and self-discipline. Finally, Openness to Experience is represented by tendencies such as having an interest in cultural events, being creative, and holding nontraditional beliefs. Convergence between the FFM and other major adult personality models has been well-documented (e.g., John and Srivastava, 1999 and Markon et al., 2005), allowing greater understanding of how various models relate to one another. However, as noted above, the field of temperament and childhood personality has been lacking similar integration. Toward this end, recent attempts have offered potential frameworks to integrate various models of childhood personality (Halverson et al., 2003 and Shiner and Caspi, 2003). The popularity of the FFM in adult personality research has contributed to a growing body of literature documenting investigations of the five factor structure in children and adolescents (e.g., Digman, 1990, Graziano and Ward, 1992 and John et al., 1994Lamb et al., 2002 and Martin et al., 1994). These studies tend to find evidence for the five factor structure, although there is some indication that Openness to Experience may not emerge until adolescence (Lamb et al., 2002). Halverson et al. (2003) took a “bottom-up” approach to the structural debate by attempting to construct a cross-cultural, cross-age measure of childhood personality. They collected parental descriptors of children (ages 3–12) across eight countries and, using a combination of rational (sorting done by focus groups) and empirical (factor analytic) techniques, derived a robust model of childhood personality. The resulting model consists of 15 “midlevel” scales that fit into a five factor structure similar to the FFM, although Openness to Experience was marked only by the single “Intellect” scale, which reflects characteristics such as precociousness, intelligence, and quickness in comprehension. Overall, the authors concluded that the results of their efforts produced a preliminary comprehensive taxonomy of childhood personality. Shiner and Caspi (2003) also proposed a preliminary comprehensive taxonomy for childhood personality, which the authors recently extended (Caspi et al., 2005 and Caspi and Shiner, 2006). They formulated their proposed taxonomy based on a review of the literature on childhood and adolescent personality, and also integrated existing and emerging work on the structure of adult personality. They initially defined a classification system consisting of four higher-order traits and eleven lower-order traits (Shiner & Caspi, 2003), recently extending the taxonomy to include a fifth higher-order trait (Caspi et al., 2005 and Caspi and Shiner, 2006). The higher-order traits map onto the five factors in the FFM. Questions remain over whether the fifth factor in the FFM, Openness to Experience, has a direct analog in childhood (Shiner, 2006) and whether it emerges later in development (Caspi & Shiner, 2006). These two recent attempts to provide a preliminary classification system for childhood and adolescent personality virtually agreed on the relevant higher-order traits. This convergence on a higher-order structure is particularly compelling given the different approaches to constructing a taxonomy that were taken by the authors. Halverson et al. (2003) employed a bottom-up scale construction approach to produce scales that were robust cross-culturally and cross-age (within childhood). Alternately, Shiner and Caspi (2003) constructed a taxonomy based on reviewing and integrating literature to date on childhood and adolescent personality. In addition, both of these works compared their proposed scales/traits with variables identified in other measures of childhood personality or temperament, allowing better understanding of how particular personality or temperament traits might be compared across studies. For example, some other models of higher-order personality factors posit three instead of five factors (e.g., Rothbart, Ahadi, Hershey, & Fisher, 2001), and may be related to the FFM in a hierarchical manner with the three factors found in these models occupying a level of the hierarchy above the five factors of the FFM, as has been demonstrated with adult samples (Markon et al., 2005). In a rigorous investigation, Markon et al. (2005) demonstrated that three-factor models (including traits roughly capturing positive emotionality, negative emotionality, and disinhibition) can be broken down hierarchically into a four-factor model (with disinhibition splitting into disagreeable disinhibition and unconscientious disinhibition). Further, at a lower level of the hierarchy, a fifth factor, roughly representing openness to experience, splits off from positive emotionality. Such an investigation has not been conducted with children, and it may well be that the hierarchical nature of higher-order personality traits differs by age. Nonetheless, this work offers some insight into potential relationships at the higher-order personality trait level between measures. Taken together, these studies provide a good starting point for discussing childhood personality within a comprehensive framework. Moreover, the framework put forth will allow for greater communication within the area of childhood personality, as well as linking the areas of childhood and adult personality (Caspi & Shiner, 2006). 1.3. Defining personality in the present review The proposed taxonomies for childhood personality (Halverson et al., 2003 and Shiner and Caspi, 2003) discussed above will provide the guiding framework for conceptualizing childhood and adolescent personality at the higher-order trait level in the present review. Thus, research on childhood and adolescent personality which utilizes other measures or variables (including relevant traits typically defined as “temperament”) will be discussed as it relates to these taxonomies. In other words, discussion through the remainder of this review will not strictly differentiate studies that refer to temperament traits versus personality traits for the purpose of maintaining a common language and integrating this body of work. In some cases, personality traits or variables that do not have a direct correlate in the proposed taxonomies will be included if deemed particularly relevant. In addition, this review will primarily focus on personality and psychopathology in children and adolescents. Thus, research on infants and toddlers will not be included. There are several reasons for making this distinction in the present review. First, as noted previously, the developmental relationships between specific temperament traits and later personality traits have not yet been fully elucidated, particularly to the extent that we do not currently have an empirical understanding of how traits in major temperament models may develop into traits in major models of personality (although this concern is somewhat informed by recent advances in extending temperament approaches into childhood and adolescence). Therefore, it is not yet clear how temperament research on infants and toddlers might be integrated with research on childhood and adolescent personality. Secondly, there is evidence to suggest that personality traits can be identified and reliably measured in early to middle childhood. Halverson et al. (2003) found that the personality variables identified in their study could be reliably measured as early as 3 years of age. Similarly, a recent meta-analysis found that rank-order consistency of personality traits increased substantially from the infancy and toddler years to the 3 to 6 year period (Roberts & DelVecchio, 2000). In addition, this meta-analysis suggested that, starting at a 3 to 6 year age range, rank-order consistency largely leveled off through childhood and adolescence (and remained in the .45 to .52 consistency range). Other research has shown that personality traits measured at age 3 were significantly linked to personality traits at age 18, suggesting that there is continuity in personality even over the range of middle childhood and adolescence (Caspi, 2000 and Caspi and Silva, 1995). It is important to interpret the magnitude of stability coefficients over longer periods of time with the understanding that while many important developmental processes and interactions with the environment also play a role in the developing personality, a significant portion of the variance (particularly relative to others in an age-specific population, i.e., rank-order consistency) appears to be stable across childhood and adolescence.