AbstractBackgroundIn Alzheimer’s disease (AD), plasma biomarkers of neurodegeneration and phosphorylated tau have initial support for their relationship to clinical diagnosis and cognitive outcomes over time (total tau [t‐tau], neurofilament light [NfL], and p‐tau). To date, only one study has assessed a network model with these data. Here, we assessed the network relations (i.e., conditional dependencies) between diagnostic variables, plasma biomarkers, neuropsychological test performance, and demographic variables using a Gaussian graphical model (GGM) with participants from the Boston University (BU) Alzheimer’s Disease Research Center (ADRC) Longitudinal Clinical Core Registry.MethodThe sample included individuals with normal cognition (n=235), MCI due to AD (n=181), and AD dementia (n=153). Participants completed a comprehensive battery of neuropsychological tests to assess global cognition, attention, executive function, episodic memory, and language abilities. Diagnoses were adjudicated during multidisciplinary diagnostic consensus conferences. Plasma samples were analyzed using the Simoa technique. Regularized GGMs were estimated with Pearson, polychoric, and polyserial correlations depending on whether the variables were continuous, ordinal, or mixed. Network stability was assessed with a non‐parametric bootstrap of standardized edge weights and a person‐dropping bootstrap of centrality measures with 10,000 samples. Latent dimensions were estimated with Markov chain Monte Carlo (MCMC) simulation.ResultAssessment of the model suggested adequate fit, χ2(184, N = 569) = 620.46, EBIC = 27,444.53, RMSEA = 0.06, TLI = 0.95. Metrics of strength, closeness centrality, and expected influence met established thresholds of stability. Among biomarkers, diagnosis was connected to p‐tau181, βz = 0.07, 95% CI [0.01, 0.12]. Assessment of network centrality suggested that p‐tau181 ranked relatively highly for betweenness centrality and expected influence and NfL ranked highly for betweenness centrality. In contrast, t‐tau was not connected to any functional or diagnostic variable and did not rank highly for centrality. The biomarker variables loaded similarly on the two latent dimensions and were clustered near consensus conference diagnosis.ConclusionUnlike t‐tau, plasma biomarkers p‐tau181 and NfL were influential as “central junctions” for other connections in a network of diagnostic, biomarker, and demographic data. The findings support recent research that posits p‐tau181 and NfL reflect distinct aspects of AD progression (respectively, AD pathology and neurodegeneration).
Read full abstract