Affiliations: [a] Department of Family Medicine, University of North Texas Health Science Center, Institute for Translational Research, Fort Worth, TX, USA
| [b] Department of Pharmacology & Neuroscience, University of North Texas Health Science Center, Institute for Translational Research, Fort Worth, TX, USA
| [c] Department of Psychology, University of North Texas, Denton, TX, USA
Correspondence:
[*]
Correspondence to: Melissa Petersen, PhD, University of North Texas Health Science Center, Department of Family Medicine, Institute for Translational Research, 3500 Camp Bowie Blvd, Fort Worth, TX 76107, USA. Tel.: +1 817 735 2959; +1 817 735 0611; E-mail: Melissa.Petersen@unthsc.edu.
Abstract: Background:Recent work has supported use of blood-based biomarkers in detection of amnestic mild cognitive impairment (MCI). Inclusion of neuropsychological measures has shown promise in enhancing utility of biomarkers to detect disease. Objective:The present study sought to develop cognitive-biomarker profiles for detection of MCI. Methods:Data were analyzed on 463 participants (normal control n = 378; MCI n = 85) from HABLE. Random forest analyses determined proteomic profile of MCI. Separate linear regression analyses determined variance accounted for by select biomarkers per neuropsychological measure. When neuropsychological measure with the least shared variance was identified, it was then combined with select biomarkers to create a biomarker-cognitive profile. Results:The biomarker-cognitive profile was 90% accurate in detecting MCI. Among amnestic MCI cases, the detection accuracy of the biomarker-cognitive profile was 92% and increased to 94% with demographic variables. Conclusion:The biomarker-cognitive profile for MCI was highly accurate in its detection with use of only five biomarkers.
