Hypofrontal activity during word retrieval in older adults: A near-infrared spectroscopy study

Abstract

The supplementary motor area (SMA) has been regarded as a third speech area. The SMA is anatomically classified into two regions, pre-SMA and SMA proper, but the functional specialization of speech production between the two regions remains unknown. Although word retrieval difficulties were often observed in older adults, there was no report as to whether the SMA would be involved in the retrieval difficulties. We focused on the SMA as a function of word production and then used near-infrared spectroscopy (NIRS) with the verbal fluency task (VFT) to explore the possible mechanism underlying the retrieval difficulties related to aging. Based on the anatomical differences within the SMA, we relied on region-of-interest (ROI) analysis to compare the brain activation patterns in the SMA during VFT between 11 healthy elder and 11 younger subjects in the situation where both groups show comparable task performance. Notably, the anterior VFT-related SMA response was more robust in the younger than in the elder group. Furthermore, anterior SMA responses in the elder group may only have a positive correlation with the VFT performance. The findings imply that anterior SMA hypoactivity in elders may cause word retrieval difficulties, while bilateral prefrontal cortices, having close connection with the pre-SMA, may contribute to the compensatory process that enables equivalent performance of the elder group with the younger one.

Introduction

The supplementary motor area (SMA) has been shown to be involved in several aspects of motor control, including movement selection, preparation, initiation, execution of complex sequences of movement, as well as motor learning (Hikosaka et al., 1996, Picard and Strick, 2003, Rao et al., 1993, Tanji and Shima, 1994). Especially, current opinions regarding the function of the SMA revolve around its role in word production of speech (Fried et al., 1991). Previously, Penfield postulated that the SMA might be a third speech area, based on the evidence of vocalization and speech arrest by direct electric stimulation of the SMA (Penfield, 1950, Penfield and Welch, 1951). A previous article claimed that unilateral SMA damage results in transient impairment of speech and that the prognosis for recovery is usually good, implying the bilateral contribution of the SMA to the generation of speech (Zentner, Hufnagel, Pechstein, Wolf, & Schramm, 1996). An functional magnetic resonance imaging (f-MRI) study with aphasic patients suggested that the most consistent compensatory activation was found in the SMA, compared with healthy control subjects (Karbe et al., 1998). The findings suggested a functional reorganization of speech by the SMA. In this regard, accumulating evidence has demonstrated that the SMA should be functionally divided into at least two distinct areas, the SMA proper, posterior to the vertical anterior commissorial (VAC) line, perpendicular to the anterior commissure (AC)—posterior commissure (PC) plane, and the pre-SMA, anterior to the VAC line (Matsuzaka et al., 1992, Tanji, 1996, Tanji and Shima, 1994). In fact, the SMA proper and pre-SMA are anatomically different. The former receives information from all components of the motor system (Jurgens, 1984), whereas the latter is densely interconnected with the prefrontal cortex (PFC) and also receives input from basal ganglia and the cerebellum, but has no connection with the motor system (Akkal, Dum, & Strick, 2007). Also, both SMA proper and pre-SMA may be involved in speech production (Fontaine et al., 2002, Tremblay and Gracco, 2009), but any functional specialization of speech production between the two regions remains unknown.

Recently, there has been much progress in the development of advanced non-invasive neuroimaging techniques, such as f-MRI and positron emission tomography (Gaillard et al., 2003) and they have become widely used for fundamental brain research in humans (Gaillard et al., 2003) and non-human primates as human models (Obayashi, 2004, Obayashi et al., 2007, Obayashi et al., 2009, Obayashi et al., 2001, Obayashi et al., 2003, Obayashi et al., 2002, Obayashi et al., 2004). Among these techniques, near-infrared spectroscopy (NIRS) can detect cortical activities by measuring the product of the optical path length for oxygenated and deoxygenated hemoglobin in the cortex. Advantages of NIRS over other techniques are considered to be its noninvasiveness, handiness, and measurability without constraint. In practical terms, NIRS can be performed with the subject sitting in an upright position in a less confined space than for f-MRI and PET. In order to measure overt speech in healthy subjects using NIRS, the verbal fluency task (VFT) seems suitable (Herrmann et al., 2003, Schecklmann et al., 2008, Tsujii et al., 2009). So far, a VFT was used for NIRS measurement in healthy subjects with probes positioned over F₃ (Herrmann et al., 2003), referring to the international 10/20 system for EEG electrode placement. Some studies with other modalities, such as f-MRI and PET, consistently revealed that diffuse frontal association areas, spreading over the dorsolateral frontal cortex, orbitofrontal area, and the SMA including pre-SMA and SMA proper, were activated during VFT (Abrahams et al., 2003, Alario et al., 2006, Chung et al., 2005, Tremblay and Gracco, 2009). Accordingly, the findings may suggest the possibility of the involvement of the SMA in VFT. At the very least, the use of VFT would be helpful for facilitating an understanding of the involvement of the SMA in language. However, there are no reports concerning whether NIRS could measure certain SMA activities during VFT.

It is noteworthy that word retrieval difficulties are frequently observed in older adults, but age-related changes in language function have received little attention (Shafto, Burke, Stamatakis, Tam, & Tyler, 2007). Phonemic VFT, especially, is reportedly subject to aging (Brickman et al., 2005), so a VFT might be suitable for exploring the mechanism underlying the effect of aging on word production. However, what happens in the aging brain in association with VFT decline has so far remained unknown. To address this issue, it was deemed useful to compare brain activation patterns during VFT between elder and younger subjects in the situation of both groups showing comparable task performance. Although there is still little evidence from NIRS studies (Herrmann et al., 2003, Kahlaoui et al., 2012, Kameyama et al., 2004), those studies consistently demonstrated that while the VFT performance of younger and older adults was comparable, VFT-related NIRS responses in the PFC were different between the two groups, as younger adults exhibited more left hemisphere activation while older ones did not show a lateralization effect. This reduction may be due to functional reorganization occurring during aging, thus resulting in compensation that enables the elderly to perform at a comparable level to the younger group. Nonetheless, some specific brain regions closely related to word retrieval difficulties should exist. Provided that the PFC is engaged in the compensatory process of word retrieval difficulty, the pre-SMA, having reciprocal connection with the bilateral PFC, would be a candidate for the specific brain region responsible for the difficulty. This speculation may be strengthened by other profiles of the SMA, such as the involvement in speech, contribution to the restoration process from aphasia, and the adjacent part of the hypo-PFC activity to which various cognitive declines are attributed. If so, the pre-SMA in elders would be less activated during VFT than that in the younger group with comparable VFT performance, and the VFT-related hypoactivity would depend on the VFT performance within the elder group. However, some reports have demonstrated that both oxy-Hb concentration and VFT performance are subject to the aging effect (Brickman et al., 2005, Hock et al., 1995), indicating that we need to apply the partial correlation coefficient of intra-group analysis to exclude the effect in the older adults.

Thus, in the present study, we tested whether the SMA responses were significantly altered during VFT, based on the extent of NIRS being sufficiently sensitive to the activities. In the situation where VFT performance was equivalent between the younger and older groups, we relied on region-of-interest (ROI) analysis to compare VFT-driven hemodynamic responses of the SMA between groups, based on the anatomical differences within the SMA. Further, we tested the correlation significance between VFT performance and the hemodynamic responses at each ROI within the older group, adjusting for aging by partial correlation analysis.