Outcome-specific value of longitudinal clinical information and model complexity for dynamic prediction of type 2 diabetes-related outcomes: a pooled landmarking study
- 1Shenzhen Hospital (Futian), Guangzhou University of Chinese Medicine, Guangdong , ,
Abstract
Objective: To evaluate the outcome-specific incremental value of longitudinal clinical history and model complexity for annual dynamic prediction of multiple type 2 diabetes-related outcomes using pooled landmarking.
Methods: We included 646 patients with complete annual records from baseline through year 3. Landmark times at years 1 and 2 were used to predict the first recorded occurrence of hypertension, coronary heart disease, diabetic kidney disease, diabetic retinopathy, or metabolic dysfunction-associated steatotic liver disease during the subsequent year. Four progressively expanded feature sets were compared using elastic-net logistic regression and Light Gradient Boosting Machine. The primary history-current estimand was a paired within-algorithm contrast with the risk set, validation splits, preprocessing, tuning, calibration, and evaluation records held constant. Internal validation used patient-level repeated nested cross-validation. We also evaluated fully nested model selection and model-development stability using 500 patient-level bootstrap replicates with out-of-bag evaluation.
Results: Pooled out-of-fold area under the receiver operating characteristic curve ranged from 0.722 to 0.952 across the five representative models. The paired history-current comparison showed higher AUC for hypertension with elastic-net (ΔAUC 0.097, 95% confidence interval 0.057 to 0.136) and LightGBM (0.118, 0.081 to 0.158), and for metabolic dysfunction-associated steatotic liver disease with elastic-net (0.022, 0.002 to 0.042) and LightGBM (0.025, 0.009 to 0.041). The other three outcomes showed no comparable improvement. Algorithm differences were modest and outcome dependent.
Conclusion: In this single-center cohort with complete annual T0–T3 follow-up, the incremental value of longitudinal information and model complexity was outcome specific. Longitudinal history showed greater benefit for selected outcomes, whereas algorithm differences were modest and outcome dependent. The smaller MASLD signal was supported primarily by the pooled L1–L2 analysis. Independent external validation is required before clinical application.
Related
- Rethinking Input Complexity in Transformer-Based Clinical Prediction: Implications for Feature Dimensionality and Sequence Length in Longitudinal EHR Data
- Health information sources and behaviour changes in a community population: a longitudinal study
- Circadian syndrome and incident multisite pain among middle-aged and older Chinese adults: component-specific and longitudinal analyses from CHARLS
- Longitudinal Drain-Fluid Proteomics Reveals Procedure-Specific Signatures of Postoperative Leakage after Pancreatic and Colorectal Surgery
- Repeated Specific Health Guidance and improvement of metabolic syndrome in middle-aged working men: a longitudinal observational study using modified Poisson regression
BibTeX
@article{fu2026outcomespecific,
title = {{Outcome-specific value of longitudinal clinical information and model complexity for dynamic prediction of type 2 diabetes-related outcomes: a pooled landmarking study}},
author = {Meiling Fu and Xiaohong Liu and Ting Shi and Yuxuan Lu and Jie Zou and Xiaowei Huang and Qingmei Pan and Yun Chen and Yinqin Zhong},
journal = {Frontiers in endocrinology},
year = {2026},
month = oct,
doi = {10.3389/fendo.2026.1964592},
url = {https://europepmc.org/article/PMC/PMC13671971},
note = {CC BY},
}
RIS
TY - JOUR
TI - Outcome-specific value of longitudinal clinical information and model complexity for dynamic prediction of type 2 diabetes-related outcomes: a pooled landmarking study
AU - Meiling Fu
AU - Xiaohong Liu
AU - Ting Shi
AU - Yuxuan Lu
AU - Jie Zou
AU - Xiaowei Huang
AU - Qingmei Pan
AU - Yun Chen
AU - Yinqin Zhong
PY - 2026
DA - 2026/10/02
JO - Frontiers in endocrinology
DO - 10.3389/fendo.2026.1964592
UR - https://europepmc.org/article/PMC/PMC13671971
AB - Objective: To evaluate the outcome-specific incremental value of longitudinal clinical history and model complexity for annual dynamic prediction of multiple type 2 diabetes-related outcomes using pooled landmarking. Methods: We included 646 patients with complete annual records from baseline through year 3. Landmark times at years 1 and 2 were used to predict the first recorded occurrence of hypertension, coronary heart disease, diabetic kidney disease, diabetic retinopathy, or metabolic dysfunction-associated steatotic liver disease during the subsequent year. Four progressively expanded feature sets were compared using elastic-net logistic regression and Light Gradient Boosting Machine. The primary history-current estimand was a paired within-algorithm contrast with the risk set, validation splits, preprocessing, tuning, calibration, and evaluation records held constant. Internal validation used patient-level repeated nested cross-validation. We also evaluated fully nested model selection and model-development stability using 500 patient-level bootstrap replicates with out-of-bag evaluation. Results: Pooled out-of-fold area under the receiver operating characteristic curve ranged from 0.722 to 0.952 across the five representative models. The paired history-current comparison showed higher AUC for hypertension with elastic-net (ΔAUC 0.097, 95% confidence interval 0.057 to 0.136) and LightGBM (0.118, 0.081 to 0.158), and for metabolic dysfunction-associated steatotic liver disease with elastic-net (0.022, 0.002 to 0.042) and LightGBM (0.025, 0.009 to 0.041). The other three outcomes showed no comparable improvement. Algorithm differences were modest and outcome dependent. Conclusion: In this single-center cohort with complete annual T0–T3 follow-up, the incremental value of longitudinal information and model complexity was outcome specific. Longitudinal history showed greater benefit for selected outcomes, whereas algorithm differences were modest and outcome dependent. The smaller MASLD signal was supported primarily by the pooled L1–L2 analysis. Independent external validation is required before clinical application.
SN - 1664-2392
ER -