Why the anterior cingulate cortex (ACC) is active in such a wide variety of cognitive and emotional tasks has become an important issue in recent research. In this issue of Brain, Critchley et al. (2003) report changes in heart rate in both cognitive and motor tasks related to the strength of activation in the ACC. They interpret their findings as supporting a view that the anterior cingulate is involved in the regulation of autonomic activity according to the current context of behavior.
At the turn of the twentieth century, Broca used the term ‘limbic’ to describe the medial convolution that borders the corpus callosum of each hemisphere. Originally, this term was merely descriptive, implying no function. Later, Broca would ascribe olfactory functions to the limbic lobe, and this suggestion dominated teaching of functional neuroanatomy well into the mid‐twentieth century despite a persuasive proposal by Papez in 1937. Based upon the reciprocal connections of the cingulate cortex with the hypothalamus, Papez proposed that the cingulate cortex was involved in the subjective experience of emotions.
In the 1940s and 1950s, evidence accumulated from animal as well as human studies of the ACC, which are well cited by Critchley et al. (2003). These studies demonstrated the involvement of the cingulate cortex in motor and autonomic control. For example, from his studies with monkeys Ward (1948) concluded, ‘It is now clear that area 24 has two main functions. It is the most powerful of the cortical suppressor areas and it is also a potent autonomic effector region’ (p. 21).
Results from modern studies of the ACC have been divided. On the one hand, results from neuroimaging studies have implicated the ACC in a host of cognitive functions (Posner and DiGirolamo, 1998). On the other, neuroimaging studies have also identified the ACC’s involvement in transient mood changes (Mayberg et al., 1999), depression and anxiety disorders (Mayberg et al., 2000; Brody et al., 2001) and the perception of pain (Rainville et al., 1997).
Given the findings that relate cognitive processes to the ACC, it is somewhat surprising that lesions to the ACC do not produce massive or consistent cognitive deficits. Early on, it was recognized that patients with lesions to the frontal lobe, including the ACC, demonstrated normal performance on a host of neuropsychological and intelligence tests (Rylander, 1947; Teuber, 1964). More recent studies have basically confirmed these early observations, but they also show that patients with ACC lesions do demonstrate performance deficits on the Stroop task and other tasks that have been shown to activate the ACC (Cohen et al., 1999; Ochsner et al., 2001). Perhaps the most consistent observation regarding changes associated with lesion of the ACC is with regard to affect. These patients have been described as apathetic and unconcerned when significant events occur, such as making mistakes (Eslinger and Damasio, 1985; Rylander, 1947).
How do we reconcile this division? One influential approach has been to divide the ACC into subregions that are separately responsible for affective and cognitive processes (Allman et al., 2001; Bush et al., 2000; Paus, 2001). However, the results presented by Critchley et al. suggest that this may not be the only approach. These authors suggest instead that cognition and behavior (i.e. motor acts) must produce an appropriate physiological state for it to be adaptive. This idea was proposed previously by Nauta (1971) and later by Damasio et al. (1990), but what is most critical is that Critchley et al. provide a clear experimental demonstration of this integration within the dorsal ACC, which has been considered the cognitive division. By demonstrating the influence of high‐level cognitive activity and simple motor acts on the autonomic nervous systems Critchley et al. may have provided an important model for studying the mechanisms by which mental processes are integrated with bodily systems.
Critchley et al.’s approach can also help us understand results that have until now been difficult to integrate. For example, an issue that is still being debated is the role of the ACC in conflict and error monitoring (Carter et al., 1998; Falkenstein et al., 2000). A way to think about the role of the ACC in both cognitive processes is that these processes produce autonomic reactions that signal the requirement for adaptive control of behaviour. Autonomic regulation would also fit the role of the ACC in pain perception. This integrative view is also consistent with evidence linking theta activity (a neurophysiological index of cognitive control or effort) with autonomic functions during sustained attention (Kubota et al., 2001). Like Critchley et al., Kubota et al. found midline theta electrophysiological responses (thought to be generated by the ACC) to be specifically related to activity of the sympathetic nervous system.
The involvement of the cingulate cortex in the regulation of autonomic processes can be understood by its connection with brainstem structures, but there are also substantial cortical connections (Allman et al., 2001) that tie ACC to lateral frontal and parietal areas. For example, it is suggested the ACC is involved in detecting when strategic control is required and that the lateral prefrontal cortex is involved in strategic control (MacDonald et al., 2000). How can this strategic control be understood with respect to autonomic control?
Some current findings may not fit quite so well within the role of the ACC as an autonomic regulator. One puzzle is that the anterior cingulate is very metabolically active at rest (Raichle et al., 2001). While this might be due to a need to monitor the sympathetic system, even during rest, it does not fit so naturally within the framework provided by the current paper. Also puzzling is the evidence that during cognitive activity, such as providing a use for a visually presented object, ACC activity begins within the first 150 ms of a task that takes ∼1100 ms (Abdullaev and Posner, 1998). Even in the area of heart rate control some puzzles remain. Although the heart rate does speed up during target processing, it also tends to slow down while waiting for targets. Another puzzle is the finding of decreased dorsal ACC activation and increased rostroventral ACC activation in induced sadness and depression (Mayberg et al., 1999). What does this mean relative to the findings by Critchley et al., particularly because both areas have been found to be correlated with autonomic activation?