The role of frontocingulate pathways in the emotion-cognition interface: Emerging clues from depression

By emphasizing nonlinear dynamics between appraisal and emotions, Lewis’s model provides a valuable platform for integrating psychological and neural perspectives on the emotion-cognition interface. In this commentary, I discuss the role of neuroscience in shaping new conceptualizations of emotion and the putative role of theta oscillation within frontocingulate pathways in depression, a syndrome in which emotioncognition relations are dysfunctional. In the target article, Lewis provides a wide-ranging and timely theoretical formulation of emotion-cognition relations. By emphasizing (a) bidirectional interactions between appraisal and emotion; (b) lower-order psychological and neural constituents underlying the emergence of emotion–appraisal processes; and (c) large-scale functional coupling through oscillatory neurophysiological mechanisms, Lewis offers a multilevel account of appraisal-emotion interactions, fostering a better integration of emotion theory and neurobiology. In this commentary, I elaborate on two important points raised in the target article. First, I emphasize how a brain-based approach to emotion and appraisal can uniquely inform and constrain theoretical models of these complex constructs. Second, I comment on Lewis’s assertion that “phase synchrony in the theta range may underpin the functional integration of systems mediating appraisal–emotion processes” (sect. 5.4). To this end, I review recent event-related potential (ERP) findings of action monitoring (Luu et al. 2004) and electroencephalographic (EEG) findings highlighting disrupted functional connectivity within frontocingulate pathways in depression (Pizzagalli et al. 2003a). With respect to brain-based approaches to emotion and appraisal, Lewis discusses definitional problems that have hindered the development of comprehensive theories of emotion. Here, I would like to emphasize two points. First, as Lewis argues, definitions of “appraisal” and “emotion” often overlap substantially, causing formidable conundrums to theoretical approaches based on the assumption that these two constructs have distinct functions and are governed by simple, linear, and unidirectional causal processes (e.g., appraisal as a temporal and causal antecedent of emotion; Roseman & Smith 2001). Second, and more important, the definitional overlap between emotion and appraisal mirrors substantial anatomical and functional overlap among brain regions subserving affective and cognitive processes (see Davidson 2003b, for an extended discussion). That is, many brain regions subserving appraisal processes also participate in emotional functions, and vice versa. This evidence forcefully contradicts assertions that affect and cognition are subserved by separate and independent neural circuits, and speaks against the notion that affect and appraisal are subcortically and cortically mediated, respectively (e.g., Panksepp 2003). As suggested by Lewis and others (e.g., Davidson 2003b; Pizzagalli et al. 2003b), emotion is not a monolithic process but comprises different subcomponents encompassing a distributed network of cortical and subcortical systems. Acknowledging empirical data consistent with this assertion (Phan et al. 2002) has important theoretical consequences, because, as appropriately stated by Lewis, “brain function prohibits any real independence between appraisal and emotion” (sect. 5). In sum, although Lewis’s overview of neural substrates underlying appraisal and emotional processes is neither comprehensive nor new, a reconceptualization of these substrates in terms of dynamic systems is indeed useful for stressing that the brain’s anatomy places important constraints upon psychological theories of emotion and its relations to cognition. Emerging brain-based approaches to the study of depression have similarly underscored not only the synergy between emotional and appraisal processes, but also the utility of a neurobiological framework to parsing the clinical heterogeneity of the disorder (Davidson et al. 2002; Pizzagalli et al. 2004). My second set of comments pertains to the hypothesis that phase synchrony in the theta range may play a critical role in the functional integration of appraisal–emotion processes. Specifically, Lewis predicts that theta synchronization across the amygdala, hippocampus, anterior cingulate (ACC), orbitofrontal (OFC), and prefrontal (PFC) cortices may “underpin the functional integration of systems mediating appraisal–emotion processes” (sect. 5.4). In humans, empirical evidence for this hypothesis is very limited, but recent findings provide promising support. First, a recent ERP study has shown that the error-related negativity (ERN) – an ERP peak occurring 50–100 msec after the commission of an error – was largely explained by transient phase-locking of midline theta activity to the error responses within distinct frontocingulate regions (Luu et al. 2004). This finding replicated and extended a prior report that error monitoring and evaluative feedback engaged dorsal and rostral ACC sources oscillating within the theta range (Luu et al. 2003). As Luu et al. (2003) proposed, these findings indicate that action regulation mediated by the ACC is associated with entrainment of frontocingulate pathways, consistent with the general framework of Lewis’s model. A second, albeit more indirect, line of evidence suggesting that large-scale corticolimbic synchronization is crucially involved in the emergence of emotion-appraisal processes can be derived from recent findings in major depression, a clinical condition in which coordination of these states is dysfunctional (Mineka et al. 2003). In a recent study, Pizzagalli et al. (2003a) found that baseline theta activity within ACC and PFC/OFC regions was functionally coupled for control, but not depressed, subjects. In healthy controls, this functional connectivity within frontocingulate pathways is in line with anatomical data suggesting that the ACC has reciprocal connections with the dorsolateral PFC and OFC (Barbas 1992; Petrides & Pandya 1999). Disrupted functional connectivity within frontocingulate networks in depression is intriguing, particularly in light of evidence reviewed in the target article and elsewhere (Bush et al. 2000) indicating that the ACC is critically implicated in monitoring conflicting response deCommentary/Lewis: Bridging emotion theory and neurobiology through dynamic systems modeling 214 BEHAVIORAL AND BRAIN SCIENCES (2005) 28:2 mands, detecting errors, and evaluating the emotional significance of events, and may thus be a site of convergence and integration between affective and cognitive processes. The fact that functional connectivity within frontocingulate pathways emerged for the theta band (6.5–8 Hz) is consistent with the hypothesis that theta may serve a gating function for the information processing flow in corticolimbic limbic regions (Vinogradova 1995; Luu et al. 2003; 2004), thereby providing the necessary neurophysiological substrates for the emergence of adaptive emotionappraisal processes, as Lewis discusses. In sum, using a theoretical framework inspired by emerging neurobiological concepts and findings, Lewis proposes a reconceptualization of emotion-cognition relations that emphasizes nonlinear interactions between their psychological and neural constituents, ultimately giving rise to a unitary phenomenon. Large-scale corticolimbic theta synchronization is proposed as a putative neurophysiological substrate giving rise to a coordinated integration of emotion and cognition. Because the strength of any theoretical account lies mainly in its predictive validity, empirical work is now needed to test hypotheses derivable from this model, including its extension to psychopathology.