Thinking of the future and past: the roles of the frontal pole and the medial temporal lobes

Abstract

Human lesion data have indicated that the frontal polar area might be critically involved in having an insight into one’s future. Retrospective memory mediated by medial temporal lobes and related structures, on the other hand, could be used to extract one’s future prospects efficiently. In the present study, we investigated the roles of these two brain structures in thinking of the future and past by using positron emission tomography (PET) and a naturalistic task setting. We measured regional cerebral blood flow (rCBF) in healthy subjects while they were talking about their future prospects or past experiences, with regard to two different temporal windows (in years or days). Many areas in the frontal and the medial temporal lobes were activated during the future and past tasks compared with a control task requiring semantic retrieval. Among these, areas in anteromedial frontal pole showed greater activation during the future tasks than during the past tasks, showing significant effect of temporal distance from the present. Most areas in the medial temporal lobes showed greater or equivalent level of activations during the future tasks compared with the past tasks. The present results suggest that thinking of the future is closely related to retrospective memory, but that specific areas in the frontal pole and the medial temporal lobes are more involved with thinking of the future than that of the past.

Introduction

Ever since the amazing case of Phineas Gage Harlow 1848, Damasio et al 1994, studies of patients with damage to specific areas in the frontal lobes have suggested that the frontal polar area, or the most anterior part of the lateral, medial, and ventral prefrontal cortices, may play significant roles in the higher order cognitive operations in humans, such as everyday planning, decision making, as well as social and moral reasoning Stuss and Benson 1987, Fuster 1989, Damasio et al 1991, Shallice and Burgess 1991a. Knowledge about the finer level of functional organization of this area is still sparse, however. This is mainly due to the difficulty with making detailed functional localization estimates from the lesion data, which are in general widely ranging within the frontal lobes, as well as the difficulty in conceptualizing the complex and abstract nature of behavioral abnormalities observed in the patients. Also, an animal model is less useful to understand the problems in the patients and their underlying neural mechanisms because the patients’ impairments tend to be limited to the everyday social activities specific to human beings. One simple concept that could be useful to grasp the nature of the disorders after damages to the human frontal pole, however, would be that they are somehow related to a dysfunction in the cognitive processing of future prospects. A simple example would be that such patients cannot think of a plan for the forthcoming weekend (Shallice and Burgess, 1991b), or a more indirect example would be that they prefer an option that will bring them an immediate reward but significant disadvantages in the future (Bechara et al., 1996). Therefore, it is suggested that the frontal polar area may play some roles in having an insight into one’s own future.

Retrospective memory, on the other hand, could be one of the cognitive resources that are supportive to having an insight into the future Tulving 1999, Burgess et al 2000, Atance and O’Neill 2001. When we make a future plan or intend to organize future behavior, we consciously or unconsciously remember our past experiences or acquired knowledge and utilize them as an effective guideline to construct ideas about the future (for a review, see Atance and O’Neill, 2001). In other words, we are unable to have a good insight into the future without reactivating past experiences or general knowledge. Since the medial temporal lobes have been thought of as essential structures for such reactivation processes of retrospective memory Scoville and Milner 1957, Nadel and Moscovitch 1997, Fujii et al 2000, it can be speculated that the medial temporal structures, along with the frontopolar areas, would be activated while one is thinking about future prospects. Considering the widely acknowledged effect of temporal gradation when amnesic patients recall autobiographical memories (for a review, see Fujii et al., 2000), it can also be assumed that temporal distance from the present might affect cognitive and brain activities related to the prospective thought about the future, although no study has yet been conducted to test such a temporal effect in the future domain.

In the present study, we investigated the physiological significance of the frontal pole and medial temporal lobes in human thoughts about the future, using a functional neuroimaging technique as well as a simple and naturalistic task procedure, i.e., subjects talking freely about their own past as well as their own future. Reliable brain activations in the anterior prefrontal cortices and/or medial temporal lobes have been observed in previous functional neuroimaging studies using such a task paradigm Andreasen et al 1995, Ryan et al 2001, Tsukiura et al 2002, although the existing results are limited only to thinking “back” to the past. In the present study, to examine the effect of temporal direction as well as temporal distance from the present on brain activity, we measured regional cerebral blood flow (rCBF) of young healthy subjects while they were talking about their ideas concerning the future or the past with respect to two temporal windows, far (several years) and near (several days). Since we intended to examine the brain activity relevant to thinking of the future and past in as natural a manner as possible, we asked the subjects to talk about any ideas with regard to the temporal windows without further restriction of its contents. This procedure also allowed us to analyze what type of responses would be produced when healthy subjects were asked to focus their attention on the far or near future. We compared the brain activity during these tasks with a semantic retrieval task, in which involvement of any thoughts related to subject’s future or past was minimized but in which making meaningful spoken output was required to the same degree.