TY  - JOUR
AU  - Jarchow, Hans
AU  - Bobrowski, Christoph
AU  - Falk, Steffi
AU  - Hermann, Andreas
AU  - Kulaga, Anton
AU  - Põder, Johann-Christian
AU  - Unfried, Maximilian
AU  - Usanov, Nikolay
AU  - Zendeh, Bijan
AU  - Kennedy, Brian K
AU  - Lobentanzer, Sebastian
AU  - Fuellen, Georg
TI  - Benchmarking large language models for personalized, biomarker-based health intervention recommendations.
JO  - npj digital medicine
VL  - 8
IS  - 1
SN  - 2398-6352
CY  - [Basingstoke]
PB  - Macmillan Publishers Limited
M1  - DZNE-2025-01210
SP  - 631
PY  - 2025
AB  - The use of large language models (LLMs) in clinical diagnostics and intervention planning is expanding, yet their utility for personalized recommendations for longevity interventions remains opaque. We extended the BioChatter framework to benchmark LLMs' ability to generate personalized longevity intervention recommendations based on biomarker profiles while adhering to key medical validation requirements. Using 25 individual profiles across three different age groups, we generated 1000 diverse test cases covering interventions such as caloric restriction, fasting and supplements. Evaluating 56000 model responses via an LLM-as-a-Judge system with clinician validated ground truths, we found that proprietary models outperformed open-source models especially in comprehensiveness. However, even with Retrieval-Augmented Generation (RAG), all models exhibited limitations in addressing key medical validation requirements, prompt stability, and handling age-related biases. Our findings highlight limited suitability of LLMs for unsupervised longevity intervention recommendations. Our open-source framework offers a foundation for advancing AI benchmarking in various medical contexts.
LB  - PUB:(DE-HGF)16
C6  - pmid:41145883
DO  - DOI:10.1038/s41746-025-01996-2
UR  - https://pub.dzne.de/record/281829
ER  -