Premium
Utility of combined plasma amyloid beta 40, amyloid beta 42, total tau, and NfL along with a measure of cognitive functioning in detecting cognitive impairment among Hispanic, Mexican Americans compared to non‐Hispanic whites
Author(s) -
Petersen Melissa,
Hall James R.,
Zhang Fan E.,
Mozdbar Sima,
Johnson Leigh Ann,
Yaffe Kristine,
Braskie Meredith N.,
King Kevin,
Toga Arthur W.,
O'Bryant Sid
Publication year - 2021
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.056344
Subject(s) - biomarker , neurodegeneration , medicine , cognition , amyloid (mycology) , amyloid beta , oncology , verbal fluency test , beta (programming language) , cognitive impairment , cognitive decline , psychology , disease , clinical psychology , neuropsychology , pathology , psychiatry , dementia , biology , biochemistry , computer science , programming language
Background The AT(N) diagnostic framework for Alzheimer’s Disease (AD) highlights the importance of specific plasma biomarkers associated with neurodegeneration (Amyloid Beta, Tau) and neuronal injury (Neurofilament light chain [Nf‐L]). Despite a wealth of research on such plasma biomarkers of AD pathology, limited work has specifically been conducted among minority populations. Our study sought to examine a combination of AD biomarkers both alone and in combination with a select measure of cognition among Hispanic, Mexican Americans (MA) and Non‐Hispanics Whites (NHW) in detecting cognitive impairment. Method Plasma samples were assayed on n=1703 participants (n= 890 MA [n=676 Normal Cognition [NC]; n=214 Cognitively Impaired [CI]]; n= 813 NHW [n=673 NC; n=140 CI]) enrolled in a study of health disparities. Plasma Amyloid Beta 40, Amyloid Beta 42, Total Tau, and Nf‐L were assayed using Simoa. A measure of verbal fluency (FAS) was utilized. The samples were randomly split (70/30) to allow for a training and test set. Random Forest (RF) models were conducted with performance on the test set reported. Result Among MA, AD biomarkers alone produced an area under the curve (AUC) of 0.50 with a sensitivity (SN) of 0.77 and specificity (SP) of 0.17. When the AD biomarkers were included along with a measure of verbal fluency (FAS)(biomarker‐cognitive profile), AUC increased to 0.63 with an SN of 0.82 and SP of 0.17. Inclusion of demographics (age, gender, education) to the biomarker‐cognitive profile increased only SP, which reached 0.50. Among NHW, AD biomarkers alone produced an AUC of 0.35 with SN of 0.88 and SP of 0.74. The combined biomarker‐cognitive profile increased AUC to 0.66 with an SN of 0.42 and SP of 0.81. The addition of demographics further increased AUC to 0.72 (SN=0.53; SP=0.88). Conclusion Detection accuracy differed by ethnicity with a broader range shown among NHW (AUCs 0.35‐0.72) compared to MA (AUCs 0.50‐0.63). Utility of AD plasma biomarkers improved with the addition of a select cognitive measure. This is consistent with prior work showing the application of a combined model (biomarker‐cognitive profile) to distinguish those with cognitive impairment. The addition of demographic factors improved the detection only for NHW.