Nonlinear biomarker interactions in conversion from mild cognitive impairment to Alzheimer's disease

Abstract

Multiple biomarkers can capture different facets of Alzheimer's disease. However, statistical models of biomarkers to predict outcomes in Alzheimer's rarely model nonlinear interactions between these measures. Here, we used Gaussian Processes to address this, modelling nonlinear interactions to predict progression from mild cognitive impairment (MCI) to Alzheimer's over 3 years, using Alzheimer's Disease Neuroimaging Initiative (ADNI) data. Measures included: demographics, APOE4 genotype, CSF (amyloid‐β42, total tau, phosphorylated tau), [18F]florbetapir, hippocampal volume and brain‐age. We examined: (a) the independent value of each biomarker; and (b) whether modelling nonlinear interactions between biomarkers improved predictions. Each measured added complementary information when predicting conversion to Alzheimer's. A linear model classifying stable from progressive MCI explained over half the variance (R2 = 0.51, p < .001); the strongest independently contributing biomarker was hippocampal volume (R2 = 0.13). When comparing sensitivity of different models to progressive MCI (independent biomarker models, additive models, nonlinear interaction models), we observed a significant improvement (p < .001) for various two‐way interaction models. The best performing model included an interaction between amyloid‐β‐PET and P‐tau, while accounting for hippocampal volume (sensitivity = 0.77, AUC = 0.826). Closely related biomarkers contributed uniquely to predict conversion to Alzheimer's. Nonlinear biomarker interactions were also implicated, and results showed that although for some patients adding additional biomarkers may add little value (i.e., when hippocampal volume is high), for others (i.e., with low hippocampal volume) further invasive and expensive examination may be warranted. Our framework enables visualisation of these interactions, in individual patient biomarker ‘space', providing information for personalised or stratified healthcare or clinical trial design.

Early MCI‐to‐AD Conversion Prediction Using Future Value Forecasting of Multimodal Features

“…To forecast future clinical changes of an MCI subject, i.e., at t +2Δ t , a reliable predictor model must capture the effects of past feature readings (at t and t +Δ t ) and project them onto the future clinical readings. Assuming longitudinal trajectories to be piecewise, we select a linear progression model for MCI-to-AD progression to avoid overfitting to a small dataset [ 41 , 42 ].…”

Brain-age predicts subsequent dementia in memory clinic patients

“…Brain-age has also been associated with subsequent dementia in observational research cohorts [32,33]. When compared with other AD neuroimaging biomarkers such as CSF-based amyloid and tau markers, or PET markers, brain-age provided an independent contribution in predicting conversion from Mild Cognitive Impairment (MCI) to AD [34]. While promising, these research studies have a key limitation; they are unrepresentative of the general population at-risk for dementia, as research participants are likely to be more highly educated, have a higher IQ, be less ethnically diverse and have fewer comorbidities than the general population [35].…”

Better Identification of Cognitive Decline With Interleukin-2 Than With Amyloid and Tau Protein Biomarkers in Amnestic Mild Cognitive Impairment

“…For example, a recent neuroimaging study reported that patients with AD could be clustered into three subtypes with distinct topographical features of cortical atrophy and tau deposition (Jeon et al, 2019). Another study used multiple biomarkers (e.g., hippocampal volume, PET amyloid deposition, and CSF tau protein) and suggest that the neurobiological processes both act independently and interact in a nonlinear fashion during progression from aMCI to AD (Popescu et al, 2020).…”