P-034Baseline structural MRI correlates of clinical measures in the Alzheimer’s disease neuroimaging initiative

The Alzheimer's Disease Neuroimaging Initiative (ADNI) was established to develop standardized imaging biomarkers of disease that are most promising as surrogate markers for disease prevention, progression and treatment. As data is becoming publicly available from this large multisite longitudinal trial it is important to validate expected relationships between clinical measures of Alzheimer's disease (AD) and imaging data. Here we present preliminary results from a subset of the initial ADNI cohort, including elderly controls (NC) and individuals with Mild Cognitive Impairment (MCI) or early AD. To evaluate baseline structural MRI correlates of clinical measures in the ADNI. ADNI data for 144 NC (mean age 75.6y; mean MMSE 29.07), 180 MCI (mean age 74.7y; mean MMSE 27.03), and 83 AD (mean age 75.6y; mean MMSE 23.51) were studied. Dual 3D T1-weighted images were downloaded, reviewed for quality, corrected for gradient nonlinearity and B1 field distortion, and averaged to improve signal-to-noise. Using semi-automated cerebral segmentation methods from FreeSurfer, regional volumes were produced for whole brain, hippocampal and ventricular regions. All measures were normalized to total intracranial volume. Statistical analyses examined group differences in terms of normalized volumes and available psychometric instruments. Across the entire cohort Mini-Mental State Examination (MMSE) scores and the Clinical Diagnostic Rating (CDR) scale global scores showed significant correlations with all brain measures (Table). Both clinical and imaging measures showed significant differences between all three diagnostic groups (p's <0.001). Detailed morphometry results are presented separately. When within-group correlations were performed, only MCI subjects showed significant correlations between MMSE and the different MRI measures, with CDR showing no significant relationships. Findings demonstrate expected relationships between diagnostic groups, global clinical assessments, and structural MRI measures, thereby validating the ADNI multi-site data collection and analysis methodology. The use of MMSE and CDR global scores are potentially limited by range restrictions in these groups. As further clinical, genetic, and biomarker data become available from the ADNI data center this dataset will be useful in evaluating questions regarding biomarker development, and the treatment and prevention of AD. Supported by: mBIRN NCRR U24-RR021382; ADNI NIH U01-AG024904.