{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,25]],"date-time":"2026-03-25T07:12:01Z","timestamp":1774422721887,"version":"3.50.1"},"reference-count":42,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,3,25]],"date-time":"2026-03-25T00:00:00Z","timestamp":1774396800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Endocrinol."],"abstract":"<jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>This study aimed to evaluate the performance of diverse open-source large language models (LLMs) in diagnosing diabetes subtypes and comorbidities from unstructured clinical text, assessing the impact of model characteristics, prompting, and language.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We conducted a retrospective analysis of 11,329 adult diabetes patients from a large Chinese tertiary center (2010\u20132020). Various open-source LLMs were tested using four prompting strategies in English and Chinese. Primary outcomes were F1-scores for multi-class diabetes subtyping and binary classification of diabetic kidney disease (DKD) and metabolic syndrome (MetS).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>LLMs demonstrated high performance in complex subtyping (peak F1 0.951) but showed limitations in rule-based DKD (F1 0.570) and MetS (F1 0.650) diagnosis. Chain-of-Thought prompting improved MetS classification but degraded DKD performance. Optimal model size was approximately 32B parameters. Notably, English prompts outperformed Chinese prompts on native Chinese text.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>Open-source LLMs exhibit strong holistic pattern recognition for complex classification but struggle with rule-based procedural reasoning. These models are promising as clinical co-pilots to augment expert decision-making rather than serving as autonomous diagnostic tools.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fendo.2026.1747468","type":"journal-article","created":{"date-parts":[[2026,3,25]],"date-time":"2026-03-25T05:43:16Z","timestamp":1774417396000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Real-world performance of open-source large language models in diabetes diagnosis"],"prefix":"10.3389","volume":"17","author":[{"given":"Shuting","family":"Yang","sequence":"first","affiliation":[{"name":"National Clinical Research Center for Metabolic Diseases, Key Laboratory of Diabetes Immunology (Central South University), Ministry of Education, and Department of Metabolism and Endocrinology, The Second Xiangya Hospital of Central South University","place":["Changsha, Hunan, China"]},{"name":"Department of Neurosurgery, The Sixth Affiliated Hospital of Sun Yat-sen University","place":["Guangzhou, China"]},{"name":"Graceland Medical Center, The Sixth Affiliated Hospital, Sun Yat-sen University","place":["Guangzhou, Guangdong, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Sujie","family":"Liu","sequence":"additional","affiliation":[{"name":"Department of Computer Science at the School of Informatics, Xiamen University","place":["Xiamen, China"]},{"name":"National Institute for Data Science in Health and Medicine, Xiamen University","place":["Xiamen, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yuxi","family":"Ma","sequence":"additional","affiliation":[{"name":"Department of Computer Science at the School of Informatics, Xiamen University","place":["Xiamen, China"]},{"name":"National Institute for Data Science in Health and Medicine, Xiamen University","place":["Xiamen, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Baowen","family":"Gai","sequence":"additional","affiliation":[{"name":"Department of Colorectal Surgery, Department of General Surgery, The Sixth Affiliated Hospital, Sun Yat-sen University","place":["Guangzhou, Guangdong, China"]},{"name":"Guangdong Provincial Key Laboratory of Colorectal and Pelvic Floor Diseases, The Sixth Affiliated Hospital, Sun Yat-sen University","place":["Guangzhou, Guangdong, China"]},{"name":"Biomedical Innovation Center, The Sixth Affiliated Hospital, Sun Yat-sen University","place":["Guangzhou, Guangdong, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Junwei","family":"Liu","sequence":"additional","affiliation":[{"name":"Guangzhou National Laboratory","place":["Guangzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Liansheng","family":"Wang","sequence":"additional","affiliation":[{"name":"Department of Computer Science at the School of Informatics, Xiamen University","place":["Xiamen, China"]},{"name":"National Institute for Data Science in Health and Medicine, Xiamen University","place":["Xiamen, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Feng","family":"Gao","sequence":"additional","affiliation":[{"name":"Department of Colorectal Surgery, Department of General Surgery, The Sixth Affiliated Hospital, Sun Yat-sen University","place":["Guangzhou, Guangdong, China"]},{"name":"Guangdong Provincial Key Laboratory of Colorectal and Pelvic Floor Diseases, The Sixth Affiliated Hospital, Sun Yat-sen University","place":["Guangzhou, Guangdong, China"]},{"name":"Biomedical Innovation Center, The Sixth Affiliated Hospital, Sun Yat-sen University","place":["Guangzhou, Guangdong, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zhiguang","family":"Zhou","sequence":"additional","affiliation":[{"name":"National Clinical Research Center for Metabolic Diseases, Key Laboratory of Diabetes Immunology (Central South University), Ministry of Education, and Department of Metabolism and Endocrinology, The Second Xiangya Hospital of Central South University","place":["Changsha, Hunan, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1965","published-online":{"date-parts":[[2026,3,25]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"891","DOI":"10.1016\/s2213-8587(18)30052-4","article-title":"Precision medicine in the management of type 2 diabetes","volume":"6","author":"Gloyn","year":"2018","journal-title":"Lancet Diabetes Endocrinol"},{"key":"B2","doi-asserted-by":"publisher","DOI":"10.1007\/s00125-019-4863-8","article-title":"Type 1 diabetes defined by severe insulin deficiency occurs after 30 years of age and is commonly treated as type 2 diabetes","volume":"62","author":"Thomas","year":"2019","journal-title":"Diabetologia"},{"key":"B3","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-023-05881-4","article-title":"Foundation models for generalist medical artificial intelligence","volume":"616","author":"Moor","year":"2023","journal-title":"Nature"},{"key":"B4","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-023-06291-2","article-title":"Large language models encode clinical knowledge","volume":"620","author":"Singhal","year":"2023","journal-title":"Nature"},{"key":"B5","doi-asserted-by":"publisher","DOI":"10.7326\/m23-2772","article-title":"Large language models in medicine: The potentials and pitfalls: A narrative review","volume":"177","author":"Omiye","year":"2024","journal-title":"Ann Intern Med"},{"key":"B6","article-title":"Llama 2: Open foundation and fine-tuned chat models","author":"Touvron","year":"2023","journal-title":"arXiv:2307.09288"},{"key":"B7","article-title":"BRIDGE: benchmarking large language models for understanding real-world clinical practice text","author":"Wu","year":"2025","journal-title":"arXiv preprint arXiv:250419467"},{"key":"B8","doi-asserted-by":"publisher","DOI":"10.2337\/dbi20-0001","article-title":"Subtypes of type 2 diabetes determined from clinical parameters","volume":"69","author":"Ahlqvist","year":"2020","journal-title":"Diabetes"},{"key":"B9","doi-asserted-by":"publisher","DOI":"10.3389\/frai.2025.1652556","article-title":"DiaGuide-LLM-Using large language models for patient-specific education and health guidance in diabetes","volume":"8","author":"Skjervold","year":"2025","journal-title":"Front Artif Intell"},{"key":"B10","doi-asserted-by":"publisher","DOI":"10.3389\/frai.2025.1658316","article-title":"Prompt engineering for accurate statistical reasoning with large language models in medical research","volume":"8","author":"Vilakati","year":"2025","journal-title":"Front Artif Intell"},{"key":"B11","doi-asserted-by":"publisher","DOI":"10.26463\/rjms.15_1_12","article-title":"Diagnosis and classification of diabetes: Standards of care in diabetes-2025","volume":"48","year":"2025","journal-title":"Diabetes Care"},{"key":"B12","doi-asserted-by":"publisher","DOI":"10.2337\/dc14-1296","article-title":"Diabetic kidney disease: a report from an ADA Consensus Conference","volume":"37","author":"Tuttle","year":"2014","journal-title":"Diabetes Care"},{"key":"B13","doi-asserted-by":"publisher","first-page":"4","DOI":"10.3760\/cma.j.issn.1674-5809.2018.01.003","article-title":"Chinese guideline for the prevention and treatment of type 2 diabetes mellitus(2017 edition)","volume":"10","author":"Society","year":"2018","journal-title":"Chin J Diabetes Mellitus"},{"key":"B14","article-title":"The Llama 3 Herd of Models","author":"Grattafiori","year":"","journal-title":"arxiv:240721783"},{"key":"B15","article-title":"Qwen2.5 Technical Report","author":"Qwen","year":"","journal-title":"arxiv:241215115"},{"key":"B16","article-title":"Qwen2 Technical Report","author":"Yang","year":"","journal-title":"arxiv:240710671"},{"key":"B17","article-title":"Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution","author":"Wang","year":"","journal-title":"arxiv:240912191"},{"key":"B18","article-title":"Qwen-VL: A versatile vision-language model for understanding, localization, text reading, and beyond","author":"Bai","year":"","journal-title":"arxiv:230812966"},{"key":"B19","article-title":"DeepSeek-R1: Incentivizing reasoning capability in LLMs via reinforcement learning","author":"DeepSeek AI","year":"","journal-title":"arxiv:250112948"},{"key":"B20","article-title":"DeepSeek-V3 Technical Report","author":"DeepSeek AI","year":"","journal-title":"arxiv:241219437"},{"key":"B21","doi-asserted-by":"publisher","first-page":"8384","DOI":"10.1038\/s41467-024-52417-z","article-title":"Towards building multilingual language model for medicine","volume":"15","author":"Qiu","year":"2024","journal-title":"Nat Commun"},{"key":"B22","article-title":"HuatuoGPT-II, One-stage Training for Medical Adaption of LLMs","author":"Chen","year":"2024"},{"key":"B23","doi-asserted-by":"publisher","DOI":"10.3390\/ai6030056","article-title":"Emotion-aware embedding fusion in large language models (Flan-T5, Llama 2, DeepSeek-R1, and ChatGPT 4) for intelligent response generation","volume":"6","author":"Rasool","year":"2025","journal-title":"AI"},{"key":"B24","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-024-03097-1","article-title":"Evaluation and mitigation of the limitations of large language models in clinical decision-making","volume":"30","author":"Hager","year":"2024","journal-title":"Nat Med"},{"key":"B25","doi-asserted-by":"publisher","first-page":"39426","DOI":"10.1038\/s41598-025-22940-0","article-title":"Limitations of large language models in clinical problem-solving arising from inflexible reasoning","volume":"15","author":"Kim","year":"2025","journal-title":"Sci Rep"},{"key":"B26","article-title":"Chain of thought prompting elicits reasoning in large language models","author":"Wei","year":"2022"},{"key":"B27","doi-asserted-by":"publisher","DOI":"10.52202\/075280-3275","article-title":"Language models don't always say what they think: Unfaithful explanations in chain-of-thought prompting","author":"Turpin","year":"2023"},{"key":"B28","doi-asserted-by":"publisher","first-page":"86","DOI":"10.1038\/s43856-025-01348-x","article-title":"Structured clinical approach to enable large language models to be used for improved clinical diagnosis and explainable reasoning","volume":"6","author":"Ayoub","year":"2026","journal-title":"Commun Med"},{"key":"B29","doi-asserted-by":"publisher","DOI":"10.3390\/bdcc8110161","article-title":"A review of large language models in healthcare: Taxonomy, threats, vulnerabilities, and framework","volume":"8","author":"Hamid","year":"2024","journal-title":"Big Data Cogn Computing"},{"key":"B30","volume-title":"Addressing hallucination in causal Q&A: The efficacy of fine-tuning over prompting in LLMs","author":"Niess","year":"2025"},{"key":"B31","doi-asserted-by":"publisher","DOI":"10.3389\/fdata.2025.1505877","article-title":"Fine-tuning or prompting on LLMs: Evaluating knowledge graph construction task","volume":"8","author":"Ghanem","year":"2025","journal-title":"Front Big Data"},{"key":"B32","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-023-02448-8","article-title":"Large language models in medicine","volume":"29","author":"Thirunavukarasu","year":"2023","journal-title":"Nat Med"},{"key":"B33","doi-asserted-by":"publisher","first-page":"84","DOI":"10.1038\/s41524-025-01564-y","article-title":"Fine-tuning large language models for domain adaptation: Exploration of training strategies, scaling, model merging and synergistic capabilities","volume":"11","author":"Lu","year":"2025","journal-title":"NPJ Comput Mater"},{"key":"B34","doi-asserted-by":"publisher","DOI":"10.52202\/075280-0441","article-title":"Qlora: Efficient finetuning of quantized llms","volume":"36","author":"Dettmers","year":"2023","journal-title":"Adv Neural Inf Process Syst"},{"key":"B35","article-title":"GPTQ: Accurate post-training quantization for generative pre-trained transformers","author":"Frantar","year":"2022"},{"key":"B36","doi-asserted-by":"publisher","DOI":"10.1145\/3699518","article-title":"Survey on knowledge distillation for large language models: Methods, evaluation, and application","volume":"16","author":"Yang","year":"2025","journal-title":"ACM Trans Intell Syst Technol"},{"key":"B37","doi-asserted-by":"publisher","first-page":"30675","DOI":"10.1101\/2024.06.23.24309365","article-title":"Distilling the knowledge from large-language model for health event prediction","volume":"14","author":"Ding","year":"2024","journal-title":"Sci Rep"},{"key":"B38","volume-title":"ChatGPT beyond English: Towards a comprehensive evaluation of large language models in multilingual learning","author":"Lai","year":"2023"},{"key":"B39","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pdig.0000022","article-title":"Sources of bias in artificial intelligence that perpetuate healthcare disparities-a global review","volume":"1","author":"Celi","year":"2022","journal-title":"PloS Digit Health"},{"key":"B40","doi-asserted-by":"publisher","DOI":"10.1001\/jamainternmed.2018.3763","article-title":"Potential biases in machine learning algorithms using electronic health record data","volume":"178","author":"Gianfrancesco","year":"2018","journal-title":"JAMA Intern Med"},{"key":"B41","doi-asserted-by":"publisher","DOI":"10.2196\/75030","article-title":"Aiding large language models using clinical scoresheets for neurobehavioral diagnostic classification from text: Algorithm development and validation","volume":"4","author":"Lin","year":"2025","journal-title":"JMIR AI"},{"key":"B42","doi-asserted-by":"publisher","DOI":"10.3390\/info16040301","article-title":"nBERT: Harnessing NLP for emotion recognition in psychotherapy to transform mental health care","volume":"16","author":"Rasool","year":"2025","journal-title":"Information"}],"container-title":["Frontiers in Endocrinology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fendo.2026.1747468\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,25]],"date-time":"2026-03-25T05:43:27Z","timestamp":1774417407000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fendo.2026.1747468\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,25]]},"references-count":42,"alternative-id":["10.3389\/fendo.2026.1747468"],"URL":"https:\/\/doi.org\/10.3389\/fendo.2026.1747468","relation":{},"ISSN":["1664-2392"],"issn-type":[{"value":"1664-2392","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,3,25]]},"article-number":"1747468"}}