Affiliations: [a] Department of Psychiatry, the University of Texas Health Science Center, San Antonio, TX, USA
| [b] Department of Medicine, the University of Texas Health Science Center, San Antonio, TX, USA
| [c] Department of Family and Community Medicine, the University of Texas Health Science Center, San Antonio, TX, USA
| [d] The Glenn Biggs Institute for Alzheimer’s and Neurodegenerative Disease, the University of Texas Health ScienceCenter, San Antonio, TX, USA
Correspondence:
[*]
Correspondence to: Donald R. Royall, Department of Psychiatry, the University of Texas Health Science Center, San Antonio, TX, USA. E-mail: royall@uthscsa.edu.
Note: [1] These authors contributed equally to this work.
Abstract: Background:We have explored dementia’s blood-based protein biomarkers in the Texas Alzheimer’s Research and Care Consortium (TARCC) study. Among them are adipokines, i.e., proteins secreted by adipose tissue some of which have been associated with cognitive impairment. Objective:To associate adipokines with dementia severity and replicate their association across cohorts and biofluids (serum /plasma). Methods:We used eight rationally chosen blood-based protein biomarkers as indicators of a latent variable, i.e., “Adipokines”. We then associated that construct with dementia severity as measured by the latent dementia-specific phenotype “δ” in structural equation models (SEM). Significant factor loadings and Adipokines’ association with δ were replicated across biofluids in the Alzheimer’s Disease Neuroimaging Initiative (ADNI). Results:Eight adipokine proteins loaded significantly on the Adipokines construct. Adipokines measured in plasma (ADNI) or serum (TARCC) explained 24 and 70% of δ’s variance, respectively. An Adipokine composite score, derived from the latent variables, rose significantly across clinical diagnoses and achieved high areas under the receiver operating characteristic curve (ROC/AUC) for discrimination of Alzheimer’s disease from normal controls (NC) or cases of mild cognitive impairment (MCI) and between NC and MCI. Conclusion:These results again suggest that SEM can be used to create latent biomarker classifiers that replicate across samples and biofluids, and that a substantial fraction of dementia’s variance is attributable to peripheral blood-based protein levels via the patterns codified in those latent constructs.
Keywords: Adipokines, adiponectin, Alzheimer’s disease, Alzheimer’s Disease Neuroimaging Initiative, cognition, dementia, g, intelligence, leptin, resistin, Texas Alzheimer’s Research and Care Consortium
