Introduction

Identifying differences in white matter (WM) between healthy aging and those suffering from neurodegeneration is vital to understanding their pathology. With studies focusing mainly on memory, cognition and perception, little is known of age-related change in language function. This study uses Tract-Based Spatial Statistics (TBSS) and focuses on language-related regions of interests (ROIs) and investigated aging conditions in the microstructural level, quantified ROIs changes, studied relations between age and semantic and phonemic judgement, and examined the impact of aging on the brain and language-related functions. Healthy aging individuals showed significant FA decreases and increases in MD, AD, and RD in multiple language related tracts along with some traditionally not related to language. These findings suggest that language functions are disrupted with aging and are not independent of other cognitive

Background

•As the aging population grows in the near future, understanding the mechanisms and process of aging has never been more important.

•In healthy aging, white matter changes occur mainly in the prefrontal regions and temporal lobe. Tract-Based Spatial Statistics (TBSS) is widely used to investigate differences in whole-brain WM structure.

Hypothesis

•Aging affects specific neural circuits in the hippocampus and medial temporal lobe.

•Diffusion tensor imaging (DTI) should show a decrease of FA and increase of MD, AD, and RD values in the WM tracts of language with age.

•Both language-related and unrelated tracts of the WM are related with semantic and phonemic judgement since language processing is not independent of other cognitive functions as many have and rely on linguistic components.

Goal

•To investigate structural deficits in language pathways and other tracts and their relationship with degraded global cognitive functions and associated linguistic impairments in healthy aging.

METHODS

Language tasks

•In the semantic assessment, participants were given three categorical terms one at a time and told to name as many semantically related nouns as possible within 30 seconds. The scores were averaged to form an overall score.

•The phonemic assessment involved three Japanese phonemes being given, with the participants being asked to say as many words starting with the given phoneme as possible within the time limit. The scores were averaged to form an overall score.

Participants

37 individuals were prospectively recruited from National Center for Geriatrics and Gerontology in Japan, including 19 healthy elders (HE) (58% male, age: M=66.58, SD=5.23) and 18 healthy young participants (HY) (67% male, age: M=22.11, SD=3.25). Subjects completed phonemic and semantic tasks to determine their language function.

Image acquisition and analysis

•Subjects were scanned on a 3T scanner (Trio Tim, Siemens). DTI was acquired using spin echo EPI sequence. T1 images were acquired using a magnetization prepared rapid gradient-echo (MPRAGE) .

•Image preprocessing was done with FSL. Affine registration corrected for eddy current distortions and motion artifacts (FMRIB). A binary brain mask was extracted (Brain Extraction Tool) and DTIfit was used to fit DT images to each voxel and yielded voxel-wise maps of FA, MD, AD and RD for each subjects.

•Voxel-wise analyses of DTI parameter maps were done with TBSS. Nonlinear transform was applied (FNIRT) to obtain FA images registered to the FMRIB58 template and the mean FA image and skeleton were created. Two-sample t-tests were done to compare the diffusion measures between the HE and HY using a permutation-based inference tool for nonparametric statistical thresholding.

20 ROIs masks based on the JHU WM tractography atlas were created using FSL to extract the values of the diffusion indices to perform Pearson’s correlation measuring the relationships between diffusion parameters and task scores

Aging effect on diffusional indices relative to HY

White matter in relation to language performance