@article{d9e37cfa985e4a379a4af3f126382009,
title = "Identifying Blood Biomarkers for Dementia Using Machine Learning Methods in the Framingham Heart Study",
abstract = "Blood biomarkers for dementia have the potential to identify preclinical disease and improve participant selection for clinical trials. Machine learning is an efficient analytical strategy to simultaneously identify multiple candidate biomarkers for dementia. We aimed to identify important candidate blood biomarkers for dementia using three machine learning models. We included 1642 (mean 69 ± 6 yr, 53% women) dementia-free Framingham Offspring Cohort participants attending examination, 7 who had available blood biomarker data. We developed three machine learning models, support vector machine (SVM), eXtreme gradient boosting of decision trees (XGB), and artificial neural network (ANN), to identify candidate biomarkers for incident dementia. Over a mean 12 ± 5 yr follow-up, 243 (14.8%) participants developed dementia. In multivariable models including all 38 available biomarkers, the XGB model demonstrated the strongest predictive accuracy for incident dementia (AUC 0.74 ± 0.01), followed by ANN (AUC 0.72 ± 0.01), and SVM (AUC 0.69 ± 0.01). Stepwise feature elimination by random sampling identified a subset of the nine most highly informative biomarkers. Machine learning models confined to these nine biomarkers showed improved model predictive accuracy for dementia (XGB, AUC 0.76 ± 0.01; ANN, AUC 0.75 ± 0.004; SVM, AUC 0.73 ± 0.01). A parsimonious panel of nine candidate biomarkers were identified which showed moderately good predictive accuracy for incident dementia, although our results require external validation.",
keywords = "biomarkers, blood biomarkers, dementia, machine learning, risk prediction",
author = "Honghuang Lin and Himali, {Jayandra J.} and Satizabal, {Claudia L.} and Beiser, {Alexa S.} and Daniel Levy and Benjamin, {Emelia J.} and Gonzales, {Mitzi M.} and Saptaparni Ghosh and Vasan, {Ramachandran S.} and Sudha Seshadri and McGrath, {Emer R.}",
note = "Funding Information: Conflicts of Interest: E.R.M. received funding from the Health Research Board of Ireland (CSF-2020-011) and the Alzheimer{\textquoteright}s Association (AACSF-18-566570) to support this work. H.L. received funding from the National Institute on Aging (1U01AG068221-01A1), the Alzheimer{\textquoteright}s Association (grant AARG-NTF-20-643020), and the American Heart Association (Grant 20SFRN35360180) to support this work, and has received consulting fees from the University of Texas San Antonio. J.J.H. has received funding support from the NIA (R01 AG062531); C.L.S. has received funding support from the Texas Alzheimer{\textquoteright}s Research and Care Consortium (2020-58-81-CR) and has held an unpaid leadership position as Programs Chair for the ISTAART Vascular Cognitive Impairment PIA. A.S.B. has received funding support as follows: NIA (R01NS017950), NIH (R01AG054076), NIH (UH2 NS100605), NIH/NHLBI (RF1AG063507), Alzheimer{\textquoteright}s Association (2018-AARG-591645), NIH/NHLBI (75N92019D00031), NIH/NIA (RF1AG059421-01), NIH/NIA (R01AG059725), NIH (R01AG062531-01A1), and Alzheimer{\textquoteright}s Association AARG-D 2020. A.S.B. has received royalties from Cengage and has held an unpaid position on the External Advisory Committee for the South Texas Alzheimer{\textquoteright}s Center; D.L. has received consulting fees from Long Island University and University of Miami; E.J.B. has received speaker honorarium from the University of Washington, University of Alabama, University of Colorado, Stanford University, University of Chicago, University of Kentucky, University of California San Francisco, and UMCG (Groningen, The Netherlands), and is partially supported by NIH R01HL092577; R.S.V. has received NIH funding support; S.S. has received funding support from NIA, NINDS, Alzheimer{\textquoteright}s Association, ADDF, and consulting fees from Biogen; H.L. had full access to all the data in the study and takes responsibility for its integrity and the data analysis. Funding Information: Funding: ERM received funding from the Health Research Board of Ireland (CSF-2020-011) and the Alzheimer{\textquoteright}s Association (AACSF-18-566570) to support this work. HL received funding from the National Institute on Aging (1U01AG068221-01A1), the Alzheimer{\textquoteright}s Association (grant AARG-NTF-20-643020), and the American Heart Association (Grant 20SFRN35360180) to support this work. The Framingham Heart Study is supported by the National Heart, Lung, and Blood Institute contracts HHSN268201500001 and 75N92019D00031. Additional support for this work was provided by NHLBI grants 1RO1 HL64753 and R01 HL076784, grants from the National Institute on Aging (R01 AG028321, R01 AG033193, R01 AG008122, U01 AG049505, R01 AG049607, R01 AG054076, R01 AG059421, R01 AG066524, U01 AG052409), and the National Institute on Neurological Disorders and Stroke (R01 NS017950 and UH2 NS100605). The MarkVCID Consortium is funded by the National Institute of Neurological Disorders and Stroke and National Institute on Aging (U24NS100591). Publisher Copyright: {\textcopyright} 2022 by the authors. Licensee MDPI, Basel, Switzerland.",
year = "2022",
month = may,
day = "1",
doi = "10.3390/cells11091506",
language = "English (US)",
volume = "11",
journal = "Cells",
issn = "2073-4409",
publisher = "MDPI Multidisciplinary Digital Publishing Institute",
number = "9",
}