Age-related changes in the neurophysiology of language in adults: relationship to regional cortical thinning and white matter microstructure

Abstract

Although reading skill remains relatively stable with advancing age in humans, neurophysiological measures suggest potential reductions in efficiency of lexical information processing. It is unclear whether these age-related changes are secondary to decreases in regional cortical thickness and/or microstructure of fiber tracts essential to language. Magnetoencephalography, volumetric MRI, and diffusion tensor imaging were performed in 10 young (18-33 years) and 10 middle-aged (42-64 years) human individuals to evaluate the spatiotemporal dynamics and structural correlates of age-related changes in lexical-semantic processing. Increasing age was associated with reduced activity in left temporal lobe regions from 250 to 350 ms and in left inferior prefrontal cortex from 350 to 450 ms (i.e., N400). Hierarchical regression indicated that age no longer predicted left inferior prefrontal activity after cortical thickness and fractional anisotropy (FA) of the uncinate fasciculus (UF) were considered. Interestingly, FA of the UF was a stronger predictor of the N400 response than cortical thickness. Age-related reductions in left-lateralization of language responses were observed between 250 and 350 ms, and were associated with left temporal thinning and frontotemporal FA reductions. N400 reductions were not associated with poorer task performance. Rather, increasing age was associated with reduction in the left prefrontal N400, which in turn was also associated with slower response time. These results reveal that changes in the neurophysiology of language occur by middle age and appear to be partially mediated by structural brain loss. These neurophysiological changes may reflect an adaptive process that ensues as communication between left perisylvian regions declines.

Figure 1.

Sagittal views of the selected fiber tracts. Individual fiber tracts are shown projected on their corresponding T1-weighted images using Tractoview software. Color-coding is included to assist with identification of the fibers in the superimposed images.

Figure 2.

Average dSPMs of cortical responses to NW minus FF for the young and middle-aged adults for four time windows of interest. Activity is displayed on the inflated lateral and ventral surface of each hemisphere. The medial view is also shown for the early latency (80–120 ms) response. Sulci and gyri are shown in dark and light gray, respectively. Color bars range from minimum (red) to maximum (yellow) normalized dipole strength and represent signal-to-noise estimates that follow a square root of F distribution. Cool color areas represent areas of FF greater than NW condition estimates and warm color areas represent NW greater than FF condition estimates.

Figure 3.

MEG NW waveforms (red) and FF waveforms (blue) for the young (thick lines) and middle-age (thin lines) groups from 0 to 600 ms within key ROIs. y-axis values reflect noise-normalized dipole strengths within an ROI. Asterisks denote time windows where group differences were detected in the normalized NW waveform analysis. Caret (∧) denotes a trend. L, left; R, right.

Figure 4.

Cortical thickness maps demonstrating the linear association between age and thickness across the cortical surface. Blue indicates areas where increasing age was associated with thinning, while red indicates areas where increasing age was associated with increased cortical thickness. Significance thresholds range from p < 0.01 (dark blue) to p < 0.0001 (cyan).