Research Article

An Interpretable Random Forest Model for Predicting Anosognosia in Preclinical Alzheimer's Disease: Findings from the A4 Study

Volume: 52 October 22, 2026
EN TR

An Interpretable Random Forest Model for Predicting Anosognosia in Preclinical Alzheimer's Disease: Findings from the A4 Study

Abstract

Anosognosia, reduced awareness of cognitive difficulties, is common across the Alzheimer's disease (AD) continuum yet remains difficult to detect before objective impairment appears; We developed an explainable machine learning model to predict a Cognitive Function Index (CFI) discrepancy–based operational indicator of anosognosia ΔCFI-defined reduced awareness) in amyloid-positive, cognitively unimpaired older adults from the A4 study. Anosognosia was defined as a negative discrepancy between participant- and study-partner-reported CFI scores (ΔCFI<0); the comparison group (ΔCFI>0) reflects hypernosognosia, over-reporting of difficulty relative to the study partner, rather than intact awareness. Participants with ΔCFI=0 or incomplete data were excluded, yielding a final sample of 1,058. A Random Forest classifier was trained on nine features with SHapley Additive exPlanations (SHAP) values computed to quantify predictor contributions at the individual level. The model achieved 65.88% accuracy, 53.45% sensitivity, 70.59% specificity. Hippocampal volume (HV) was the highest predictor by permutation importance, followed by age and study-partner cohabitation status; FCSRT and PACC contributed meaningfully, while MMSE, GDS, and education showed near-zero importance. SHAP analyses confirmed that lower HV, older age, and lower cognitive scores jointly drove predictions of anosognosia, while misclassifications arose from competing demographic and neuroimaging signals. These findings demonstrate that an accessible combination of clinical and neuroimaging variables can identify meaningful predictors of anosognosia in preclinical AD. HV was the leading predictor, consistent with the established role of medial temporal lobe integrity in self-monitoring. The asymmetry among sensitivity, specificity, and positive predictive value underscores the need to incorporate additional biomarkers, particularly tau pathology, functional connectivity, and longitudinal data.

Keywords

References

  1. 1. Tagai K, Nagata T, Shinagawa S, Shigeta M. Anosognosia in patients with Alzheimer's disease: current perspectives. Psychogeriatrics. 2020-05-01 2020;20(3):345-352. doi:10.1111/psyg.12507
  2. 2. Starkstein SE, Jorge R, Mizrahi R, Robinson RG. A diagnostic formulation for anosognosia in Alzheimer's disease. J Neurol Neurosurg Psychiatry. Jun 2006;77(6):719-25. doi:10.1136/jnnp.2005.085373
  3. 3. Starkstein SE. Anosognosia in Alzheimer's disease: diagnosis, frequency, mechanism and clinical correlates. Cortex. Dec 2014;61:64-73. doi:10.1016/j.cortex.2014.07.019
  4. 4. Flores-Vázquez JF, Ramírez-García G, Marrufo-Meléndez OR, et al. Anosognosia in Amnestic Mild Cognitive Impairment Is Related to Diminished Hippocampal Volume Comparable to Alzheimer’s Disease Dementia: Preliminary MRI Findings. Frontiers in Aging Neuroscience. 2021-10-28 2021;13doi:10.3389/fnagi.2021.739422
  5. 5. López-Martos D, Sánchez-Benavides G, Grau-Rivera O, et al. Clinical and Pathological Progression of Awareness Trajectories in Preclinical Alzheimer's Disease. medRxiv. Feb 18 2026;doi:10.64898/2026.02.16.26346402
  6. 6. Tondelli M, Ballotta D, Maramotti R, et al. Resting-state networks and anosognosia in Alzheimer’s disease. Frontiers in Aging Neuroscience. 2024-06-05 2024;16doi:10.3389/fnagi.2024.1415994
  7. 7. Nallapu BT, Petersen KK, Qian T, et al. A Machine Learning Approach to Predict Cognitive Decline in Alzheimer Disease Clinical Trials. Neurology. Apr 22 2025;104(8):e213490. doi:10.1212/WNL.0000000000213490
  8. 8. Lundberg SM, Lee S-I. A unified approach to interpreting model predictions. Advances in neural information processing systems. 2017;30

Details

Primary Language

English

Subjects

Neurology and Neuromuscular Diseases, Neurosciences (Other), Public Health (Other)

Journal Section

Research Article

Publication Date

October 22, 2026

Submission Date

June 9, 2026

Acceptance Date

October 8, 2026

Published in Issue

Year 2026 Volume: 52

APA
Demirsoy, I., & Ceren Akgör, M. (2026). An Interpretable Random Forest Model for Predicting Anosognosia in Preclinical Alzheimer’s Disease: Findings from the A4 Study. Journal of Uludağ University Medical Faculty, 52. https://doi.org/10.32708/uutfd.1967176
AMA
1.Demirsoy I, Ceren Akgör M. An Interpretable Random Forest Model for Predicting Anosognosia in Preclinical Alzheimer’s Disease: Findings from the A4 Study. J Uludag Univ Med Fac. 2026;52. doi:10.32708/uutfd.1967176
Chicago
Demirsoy, Idris, and Merve Ceren Akgör. 2026. “An Interpretable Random Forest Model for Predicting Anosognosia in Preclinical Alzheimer’s Disease: Findings from the A4 Study”. Journal of Uludağ University Medical Faculty 52 (October). https://doi.org/10.32708/uutfd.1967176.
EndNote
Demirsoy I, Ceren Akgör M (October 1, 2026) An Interpretable Random Forest Model for Predicting Anosognosia in Preclinical Alzheimer’s Disease: Findings from the A4 Study. Journal of Uludağ University Medical Faculty 52
IEEE
[1]I. Demirsoy and M. Ceren Akgör, “An Interpretable Random Forest Model for Predicting Anosognosia in Preclinical Alzheimer’s Disease: Findings from the A4 Study”, J Uludag Univ Med Fac, vol. 52, Oct. 2026, doi: 10.32708/uutfd.1967176.
ISNAD
Demirsoy, Idris - Ceren Akgör, Merve. “An Interpretable Random Forest Model for Predicting Anosognosia in Preclinical Alzheimer’s Disease: Findings from the A4 Study”. Journal of Uludağ University Medical Faculty 52 (October 1, 2026). https://doi.org/10.32708/uutfd.1967176.
JAMA
1.Demirsoy I, Ceren Akgör M. An Interpretable Random Forest Model for Predicting Anosognosia in Preclinical Alzheimer’s Disease: Findings from the A4 Study. J Uludag Univ Med Fac. 2026;52. doi:10.32708/uutfd.1967176.
MLA
Demirsoy, Idris, and Merve Ceren Akgör. “An Interpretable Random Forest Model for Predicting Anosognosia in Preclinical Alzheimer’s Disease: Findings from the A4 Study”. Journal of Uludağ University Medical Faculty, vol. 52, Oct. 2026, doi:10.32708/uutfd.1967176.
Vancouver
1.Idris Demirsoy, Merve Ceren Akgör. An Interpretable Random Forest Model for Predicting Anosognosia in Preclinical Alzheimer’s Disease: Findings from the A4 Study. J Uludag Univ Med Fac. 2026 Oct. 1;52. doi:10.32708/uutfd.1967176

ISSN: 1300-414X, e-ISSN: 2645-9027

Creative Commons License
Journal of Uludag University Medical Faculty is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
2023