With increasing life expectancy in many societies, cognitive decline accompanying aging or neurodegenerative disease has become a growing problem. Changes in brain structures have been associated with cognitive decline in healthy older adults (Henneman et al., 2009) and may precede the appearance of clinical symptoms of dementia by several years (Carlson et al., 2008). Despite the burgeoning elderly population in Asia, the bulk of published research on ageing humans comes from the West. The present study seeks to provide an East Asian perspective on longitudinal brain aging in a cohort of relatively healthy older adults. Here, we analyzed structural brain MRI and neuropsychological data collected over the course of 8 years to quantify annual changes in brain structures and functions in 5 cognitive domains and to investigate the relationship between these brain and cognition trends.The current sample comprised 111 relatively healthy older adults (mean age=67.1 years; SD=6.3 years) with no known active major medical conditions other than treated, uncomplicated hypertension (33.3% at baseline) or diabetes mellitus (9.9% at baseline). Performance in 5 cognitive domains were assessed: speed of processing (Trial-Making Test-A, the Symbol Search Task, and Symbol-Digit Modalities Test), executive function (Categorical Verbal Fluency test, Design Fluency test, and Trial Making Test B), attention (Digit Span Task and Spatial Span task), verbal memory (Rey Auditory Verbal Learning Test), and visuo-spatial memory (Visual Paired Associates test). Individual test scores from the subsequent 3 follow-up visits were first z-transformed with reference to baseline, and then converted to T-scores (group mean of T score at baseline = 50; SD = 10). For each time point, a composite score was derived for each of the five cognitive domains by averaging the T-scores of the respective tasks. High-resolution images of the brain were acquired using a T1-weighted MEMPRAGE sequence with a 3T Siemens Tim Trio system (Siemens, Erlangen, Germany). There were 192 sagittal slices with the following scanning parameters: TR=2530ms, TI=1200ms, FA=7°, FOV=256×256 mm, isotropic voxel dimensions of 1.0 mm. Automated measurements of brain volumes were standardized across phases and performed using FreeSurfer 5.1.0 (http://surfer.nmr.mgh.harvard.edu/). Here, we report findings regarding total cerebral, gray and white matter, hippocampal, and ventricular volumes. Estimated total intracranial volume (eTIV; Buckner et al., 2004) was used as a covariate in all statistical analyses involving brain variables to compensate for inter-individual differences in head size. We used mixed-effects model to quantify the annual change in brain volume and cognitive performance at both a group and an individual levels. Partial correlations were performed to investigate the relationships between longitudinal changes in cognitive function and brain after controlling for eTIV, age, BMI, education, gender, and blood pressure (systolic and diastolic) at baseline.All brain structures investigated showed significant annual volumetric decline (from -0.53% to -0.94%/year), and the ventricles showed significant annual expansion at 3.56%/year (Table 1). These annual changes were similar to those reported in other longitudinal studies examining healthy aging (Jack et al., 2005). In contrast, speed of processing was the only cognitive domain that revealed significant decline associated with aging (mean=-0.92%/year), while performance in the other four cognitive domains was relatively preserved with increasing age (Table 1), corroborating the work of others documenting strong age effects on speed of processing (Salthouse, 2009). Inter-individuals differences in brain and cognitive changes over time were observed (Figures 1 and 2). Associations between longitudinal changes in brain structures and cognitive performance are summarized in Table 2. After adjusting for eTIV, age, BMI, education, gender, and blood pressure, greater total cerebral atrophy was significantly associated with faster decline in verbal memory (partial r=0.24, p=0.02; Figure 3A), likely driven by the latter’s positive association with white matter atrophy (partial r=0.20, p<0.05, Figure 3B). This association is consistent with an early imaging study which linked white matter lesions with poorer recall on word lists in older adults (Breteler et al., 1994), and more recent work which revealed that greater disruption of white matter pathways predicted MCI status (Delano-Wood et al., 2012). Hippocampal atrophy and ventricular expansion were associated with poorer verbal memory (partial r=0.21; p=0.03, Figure 3C, partial r=-0.22; p=0.03, Figure 3D) and executive function (partial r=0.22, p=0.03, Figure 3E, partial r=-0.23, p=0.02, Figure 3F). Similar associations have been documented for western populations (Murphy et al., 2010; Breteler et al., 1994). Using a cohort of relatively healthy Chinese older adults, we have quantified the longitudinal changes in brain structures and cognitive functions, and provided detailed characterization of the relationships between these brain and cognitive changes over an 8-year period.