Few tasks are more daunting than trying to understand pathophysiology of developmental psychopathology. When considering scientists’ successes for other complex medical problems, prior attempts to understand pathophysiology generated major insights when scientists integrated perspectives from diverse fields. Throughout history, for example, reduced morbidity from infection followed when researchers melded insights from pharmacology, microbiology, epidemiology, and even, most recently, molecular genetics. In this issue, Gao and colleagues provide a glimpse of how developmental psychopathology researchers might use a comparable, integrative approach to answer similarly complex questions: How, exactly, do mental disorders emerge? More specifically, how can we best use insights from neuroscience to understand the complex processes that lead to their initiation and maintenance?

Research on aggression is beginning to seek answers to these questions, and these efforts suggest one way forward that might broadly apply to other forms of psychopathology. Problems with aggression are among the most concerning for both clinicians and families, given the impact on society. Early-emerging aggression predicts trajectories of later problems in various domains. Recent interest focuses on factors that modulate these early-emerging trajectories. This interest focuses specifically on perturbed emotional processes, where some processes appear to predict increased aggression (Leibenluft, Blair, Charney, & Pine, 2003) and others predict the reverse (Pine, Cohen, Cohen, & Brook, 2000). While not all findings agree, one of the most consistent observations is that high levels of fear, assessed at one point in time, predict reductions in aggression, assessed later in development.

Against this backdrop, Gao and colleagues generate insights on possible neural mediators of these relationships. They followed approximately 1,500 children from ages 3 to 8 using clinically derived measures of aggression, fear, and related constructs. Novel insights emerged through their assessment of an important psychophysiological indicator of stimulus-reinforcement (S-R) learning, termed ‘fear conditioning’, through which organisms learn about relationships among neutral and aversive stimuli. Gao and colleagues show that aggressive children manifest a reduced capacity for fear conditioning. On the surface, when considered in the context of prior work consistently linking low levels of fear to high levels of aggression, these new findings might appear unsurprising or perhaps even relatively mundane. However, viewed from other perspectives, they teach two important lessons.

The first major lesson reinforces ideas seen in other work over the past few decades, albeit in diverse ways. Namely, attempts to understand developmental change and constancy in complex, context-dependent, and multi-determined behavioral phenomena benefit greatly from a longitudinal perspective. Particular advantages accrue from research that conducts repeated assessments of risk factors and associated forms of pathological behavior (Kraemer et al., 2000). Thus, when trying to understand the origins of a behavior, repeatedly assessing factors influencing the path leading up to the behavior often teaches us much more about pathophysiology than merely assessing, at one point in time, factors manifesting contemporaneous with the behavior. In Gao et al., the association between aggression and fear conditioning, assessed at any one point in time, appears weak; the association manifests more clearly when research charts over time developmental changes in fear conditioning.

The second major lesson is more novel and therefore perhaps less obvious than the first. One major hurdle for research on pathophysiology relates to the difficulty in bridging clinical and neuroscience perspectives (Hyman, 2007). While studies attempting to forge such bridges in children are especially important, they also appear particularly difficult to perform. It is hard enough to relate clinically relevant behaviors in adults to measures of brain function; it is harder still when examining children, whose behavior changes quite dramatically in the short term, through the influence of context, and in the long term, through the influence of development. It might seem nearly impossible when trying to understand processes, such as fear conditioning, which, by their very nature, must be studied by exposing children to aversive experiences, even if only relatively mild ones, which children would be expected to avoid. This is the challenge embraced by Gao and colleagues.

When striving for basic-clinical integration, considerable advantages accrue by focusing on behaviors where the mediating brain circuitry appears similar in humans and other species. Mental health researchers possess far too few examples of clinically relevant brain–behavior associations that are strongly conserved across mammals. This presumably occurs when a behavior and associated brain-circuit configuration provided a selective advantage to a range of mammals’ common ancestors. In this situation, basic researchers can study brain–behavior relationships through invasive experiments in model brain systems, and clinical researchers can then extend these experiments, using less invasive approaches in patients, with some confidence regarding the applicability of the basic work for clinical investigations. Work in other areas of medicine demonstrates the salutatory advantage of this basic-clinical integrative approach. For research on pathological behavior, the circuitry of fear conditioning represents one of the best clinically relevant examples of strong cross-species conservation. Therefore, when Gao and colleagues link fear conditioning to aggression, they build a novel, particularly strong and important basic-clinical bridge. Viewed from this perspective, the second major lesson from Gao and colleagues reflects the importance of bridging work in neuroscience and developmental psychopathology in the service of bringing a deeper focus to ongoing research.

Both lessons lay the groundwork for other work. For example, Gao and colleagues focus on one particular measure of fear conditioning, as reflected in electrodermal activity. However, studies in rodents suggest that different physiological or behavioral indicators of conditioning can occasionally generate different conclusions about the role of fear conditioning in behavior. Therefore, future studies might collect more diverse indicators of fear and aggression than included in Gao et al., each assessed repeatedly over time. Such work then might consider whether one or another specific indicator generates concordant or discordant conclusions in terms of relationships among conditioning, clinical measures of fear, and aggression trajectories.

Other potential directions also emerge. Only indirect insights on neural correlates arise in Gao et al., since no direct measures of brain function were acquired. Brain imaging implicates a circuit encompassing the amygdala and ventral prefrontal cortex in clinically related constructs, including fear conditioning, anxiety disorders, and aggression. Future studies might use brain imaging to chart how brain function predicts each of these constructs. Most importantly, studies of amygdala-frontal function in the rodent generate mechanistic-level insights on underlying mediators of behavioral trajectories, information particularly valuable when trying to discover novel therapies (Pine, Helfinstein, Bar-Haim, Nelson, & Fox, 2009). Therefore, novel treatments for aggression might emerge through extensions of research on fear conditioning and its experimental modulation through manipulations of either experience or neurochemistry. As in other work on clinical-neuroscience integration, the generation of novel ideas about therapeutics emerging in Gao et al. probably represents the most significant, immediate, clinically relevant direction for future research.

The deepening complexity in pathophysiology research conducted over the past decade generates as many questions as it does answers. A good deal of this complexity arises from the intricate functional architecture of the brain, coupled with the fact that mental illnesses evolve in a developmental context, presumably reflecting developmental change in this already-complex architecture. Studies that confirm the importance of key principles offer guides through the complex morass of accumulating data, and so emerge valuable insights from the work presented by Gao and colleagues. By studying longitudinal changes in fear conditioning, a core, evolutionarily conserved brain function, this work illustrates the importance of two key study-design features: the charting of risk-factor/behavior associations through repeated, longitudinal assessments and the importance of clinical-neuroscience integration.

The opinions and assertions contained in this paper are the private views of the author and are not construed as official or as reflecting the views of the NIMH or the Department of Health and Human Services.


  1. Top of page
  2. References