Table 2 Mixed effects models of worsening in the neuropsychological test performances over time by AD-specific atrophy similarity in patients with aMCI and AD.

From: Machine Learning-based Individual Assessment of Cortical Atrophy Pattern in Alzheimer’s Disease Spectrum: Development of the Classifier and Longitudinal Evaluation

AD-specific atrophy similarity by time

Patients with aMCI

Patients with AD

Estimate

SE

p-value

Estimate

SE

p-value

Neuropsychological tests

 Attention

−0.018

0.030

0.563

−0.121

0.041

0.003

 Language

−0.112

0.036

0.030

−0.344

0.078

<0.001

 Visuospatial

−0.032

0.120

0.791

−0.180

0.186

0.338

 Memory

−0.370

0.276

0.186

−0.438

0.213

0.044

 Frontal/executive

−0.041

0.168

0.809

−0.542

0.169

0.002

 SNSB-D total

−0.107

0.350

0.048

−1.604

0.464

0.001

MMSE

−0.064

0.035

0.024

−0.263

0.077

0.001

CDR

0.003

0.003

0.390

0.027

0.010

0.013

CDR-SB

0.042

0.020

0.003

0.145

0.044

0.002

  1. Linear mixed effects model were performed using AD-specific atrophy similarity, time, and the interaction term between AD-specific atrophy similarity and time (AD-specific atrophy similarity by time) as fixed effects and patient as random effect. AD-specific atrophy similarity was computed using w-score based on age and education.
  2. Abbreviations: aMCI = amnestic mild cognitive impairment; AD = Alzheimer’s disease; SE = standard error; SNSB-D = Seoul Neuropsychological Screening Battery-Dementia version; MMSE = mini-mental state examination; CDR = Clinical Dementia Rating; CDR-SB = Clinical Dementia Rating sum of boxes.