CLINICAL FEATURES AND IMAGING ANALYSIS OF 38 CASES OF BEHAVIORAL VARIANT FRONTOTEMPORAL DEMENTIA

Abstract

monitoring of Alzheimer’s disease (AD) patients in clinical trials. It is easy to use, inexpensive and quantifies the cognitive impairment of subjects. The objective was to adapt and validate ADAS-Cog in Colombian populations. Methods: The ADAS-Cog was translate from English and adjusted for trans-linguistic and cross-cultural equivalences. For validation people with AD (Sporadic and AD E280A mutation), minimal cognitive impairment (MCI) and normal controls were tested. Content validity was established through a factor analysis by the principal components methods. Construct validity was determined by comparisons between the 3 groups with non-parametric statistics. Concurrent criterion validity was established comparing with MoCA and GDS, using Spearman correlations. Internal consistency was evaluated with the Cronbach’s alpha. R eproducibility was tested with intraclass correlation coefficient (ICC), and the accuracy was established with ROC curves. Results: We performed a translation from the original version, then it was discussed and by consensus the Colombian Spanish version was reached. We carried a back-translation to assure that our version did not lose its psychometric properties. Then, we did a pilot study for adjustment. Then we performed the necessary changes and obtained the final version of the test. Between 2010 and 2013 we evaluated 378 subjects, 132 normal controls, 127 with MCI (16 E280A) and 119 with AD (41 E280A). Factor analysis showed that 3 main factors explained 70.3% of the variance; Sphericity was 0.001 and KMO index was 0.936. Construct validity showed that all variables and the total score were significant for different in all groups. Concurrent validity revealed a good correlation with MoCA (-0.811) and GDS (0.892). The alpha for internal consistency was 0.916, a result that implies a slight over correlation. Reproducibility was performed an average of 12.2 6 7.7 months later, and showed an ICC of 0.91 (CI95%:0.76-0.97). Accuracy presented an AUC of 0.947(CI95% 0.923-0.971). Conclusions: We adapted and validated the ADAS-Cog for its use in our region, it will facilitate international comparisons and participation in multicenter clinical trials. It has the psychometric properties of the original test, and quantifies AD regardless of its genetic background, or presentation.