Journal Article

DKFZ-2026-00964

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Prediction of cognitive test scores: a comparison of brain structure, health, demographic, and cognitive data across adulthood.

Mendl-Heinisch, C. ; Bittner, N. ; Miller, T. ; Dellani, P. ; Bamberg, F. ; Berger, K. ; Bohmann, P. ; Decker, J. A. ; Flöel, A. ; Greiser, K. H.DKFZ* ; Harries, M. ; Kapar, J. ; Keil, T. ; Klett-Tammen, C. J. ; Krist, L. ; Kröncke, T. ; Leitzmann, M. ; Niendorf, T. ; Peters, A. ; Pischon, T. ; Riedel, O. ; Ringhof, S. ; Schlett, C. L. ; Schulze, M. B. ; Wielpütz, M. O. ; Wirkner, K. ; Caspers, S. ; Jockwitz, C.

2026
Springer International Publishing [Cham]

Abstract: Cognitive performance prediction may help identify early cognitive decline. However, the heterogeneity of research findings impedes the identification of key predictors. This study used 21,877 participants (25-74 years) from the German National Cohort (NAKO Gesundheitsstudie, NAKO) to systematically predict cognitive test scores based on brain structure, demographic, health-related, and cognitive data. Importantly, validation analyses were performed across study sites and external samples (1000BRAINS). Higher predictability was observed in the total sample compared to age-specific subgroups (10% difference in explained variance). Demographic (e.g. age) and cognitive data (e.g. memory) outperformed brain structure (e.g. grey matter volume) and health-related data (e.g. hypertension). Cognitive tests were differentially predictable, most evident between episodic memory and motor speed (R2 ≤ 0.32 versus R2 ≤ 0.18). Differences in predictability between age groups finally highlight the importance of comparing prediction outcomes between adult lifespan and age-specific groups to elucidate general and age-sensitive predictors of cognitive test scores.

Keyword(s): Age decades ; Brain structure ; Cognitive functions ; Demographic ; Health-related ; Machine learning analyses ; Prediction

Note: epub

---

Contributing Institute(s):

1. Epidemiologie von Krebs (C020)

Research Program(s):

1. 313 - Krebsrisikofaktoren und Prävention (POF4-313) (POF4-313)

Appears in the scientific report 2026

---

Database coverage:
; BIOSIS Previews ; Biological Abstracts ; Clarivate Analytics Master Journal List ; DEAL Springer ; DEAL Springer ; Ebsco Academic Search ; Essential Science Indicators ; IF >= 5 ; JCR ; SCOPUS ; Science Citation Index Expanded ; Web of Science Core Collection

---