Heterogeneity in the treatment effect of sodium-glucose cotransporter-2 inhibitor on chronic kidney disease risk reduction: a prevalent new user cohort study applying double/debiased machine learning.
8 affiliations
- 1Division of Endocrinology and Metabolism, Department of Medicine, Faculty of Medicine, Siriraj Hospital, Mahidol University, Bangkok, Thailand.
- 2Department of Clinical Epidemiology and Biostatistics, Faculty of Medicine Ramathibodi Hospital, Mahidol University, Mahidol University 4Th Floor Sukho Place Building 218/11 Sukhothai Road, Dusit, Bangkok, 10300, Thailand. [email protected].
- 3Research Affairs, Faculty of Medicine, Chulalongkorn University 10, Floor Mongkut-Phetcharat Research and Service Innovation Building, King Chulalongkorn Memorial Hospital, 1873 Ratchadamri Road, Pathum Wan, Bangkok, 10330, Thailand. [email protected].
- 4Department of Clinical Epidemiology and Biostatistics, Faculty of Medicine Ramathibodi Hospital, Mahidol University, Mahidol University 4Th Floor Sukho Place Building 218/11 Sukhothai Road, Dusit, Bangkok, 10300, Thailand.
- 5Department of Information Technology, Faculty of Medicine Ramathibodi Hospital, Mahidol University, Bangkok, Thailand.
- 6Siriraj Informatics and Data Innovation Center, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand.
- 7School of Medicine and Public Health, and Hunter Medical Research Institute, University of Newcastle, New Lambton, NSW, Australia.
- 8Centre for Public Health, School of Medicine, Dentistry and Biomedical Sciences, Queen's University, Belfast, UK.
Abstract
Background
Despite strong evidence supporting the benefits of sodium-glucose cotransporter-2 inhibitor (SGLT2i) for managing type 2 diabetes, heterogeneity of treatment response is rarely assessed. This study aims to demonstrate the heterogeneity effects of SGLT2i for preventing chronic kidney disease (CKD) via a causal machine learning model.Methods
A retrospective cohort study was conducted. Data from 2,280 type 2 diabetes patients were used to develop a double/debiased machine learning (DDML) model for CKD prediction; a conventional treatment effect (TE) model was also constructed for comparison. Model performance was evaluated using area under the receiver operating characteristic curve, and calibration plot. Subsequently, average treatment effects (ATEs) and conditional average treatment effects (CATEs) of SGLT2i versus non-SGLT2i were estimated and compared among models.Results
A neural network (NN) was selected for the DDML framework for CKD prediction. Although the conventional TE model demonstrated slightly better discrimination and calibration, the DDML-NN produced comparable results. The ATEs (95% confidence interval) estimated by DDML-NN and conventional TE models were -0.0399 (-0.0600, -0.0199) and - 0.0424 (- 0.0616, - 0.0232), respectively. Corresponding CATE ranges were -0.3945 to 0.2495 and - 0.6968 to 0.0026. Patients gaining higher benefits from SGLT2i trended to be older, male, and overweight, compared to who were benefited less from SGLT2i.Conclusion
The protective effect of SGLT2i against CKD was indicated, albeit heterogeneous. Counterfactual models including DDML-NN and conventional TE models demonstrated similar performance. However, data from a larger cohort are still required to draw a firm conclusion.
Related
- Diagnostic challenges of masked arginine vasopressin deficiency in neurosarcoidosis complicated by sodium–glucose cotransporter 2 inhibitor use
- Incidence and predictors of urinary tract infections after sodium-glucose cotransporter 2 inhibitor initiation in older adults hospitalized for heart failure
- County-Level Heterogeneity in Opioid Harm Reduction and Treatment Effects: A Simulation Modeling Analysis
- Telomere length and prevalent chronic lung disease in middle-aged and older adults: socioeconomic heterogeneity and early-life tobacco exposure in the health and retirement study
- Sodium-glucose cotransporter-2 inhibitor-associated euglycemic diabetic ketoacidosis in an adolescent with type 2 diabetes.
BibTeX
@article{kongmalai2026heterogeneity,
title = {{Heterogeneity in the treatment effect of sodium-glucose cotransporter-2 inhibitor on chronic kidney disease risk reduction: a prevalent new user cohort study applying double/debiased machine learning.}},
author = {Tanawan Kongmalai and Panu Looareesuwan and Amarit Tansawet and Sukanya Siriyotha and Patratorn Kunakorntham and Burin Boonwatcharapai and John Attia and Gareth J McKay and Ammarin Thakkinstian},
journal = {Diabetology \& metabolic syndrome},
year = {2026},
month = jul,
doi = {10.1186/s13098-026-02249-2},
url = {https://europepmc.org/article/MED/42908014},
note = {CC BY-NC-ND},
}
RIS
TY - JOUR
TI - Heterogeneity in the treatment effect of sodium-glucose cotransporter-2 inhibitor on chronic kidney disease risk reduction: a prevalent new user cohort study applying double/debiased machine learning.
AU - Tanawan Kongmalai
AU - Panu Looareesuwan
AU - Amarit Tansawet
AU - Sukanya Siriyotha
AU - Patratorn Kunakorntham
AU - Burin Boonwatcharapai
AU - John Attia
AU - Gareth J McKay
AU - Ammarin Thakkinstian
PY - 2026
DA - 2026/07/01
JO - Diabetology & metabolic syndrome
DO - 10.1186/s13098-026-02249-2
UR - https://europepmc.org/article/MED/42908014
AB - Background Despite strong evidence supporting the benefits of sodium-glucose cotransporter-2 inhibitor (SGLT2i) for managing type 2 diabetes, heterogeneity of treatment response is rarely assessed. This study aims to demonstrate the heterogeneity effects of SGLT2i for preventing chronic kidney disease (CKD) via a causal machine learning model.Methods A retrospective cohort study was conducted. Data from 2,280 type 2 diabetes patients were used to develop a double/debiased machine learning (DDML) model for CKD prediction; a conventional treatment effect (TE) model was also constructed for comparison. Model performance was evaluated using area under the receiver operating characteristic curve, and calibration plot. Subsequently, average treatment effects (ATEs) and conditional average treatment effects (CATEs) of SGLT2i versus non-SGLT2i were estimated and compared among models.Results A neural network (NN) was selected for the DDML framework for CKD prediction. Although the conventional TE model demonstrated slightly better discrimination and calibration, the DDML-NN produced comparable results. The ATEs (95% confidence interval) estimated by DDML-NN and conventional TE models were -0.0399 (-0.0600, -0.0199) and - 0.0424 (- 0.0616, - 0.0232), respectively. Corresponding CATE ranges were -0.3945 to 0.2495 and - 0.6968 to 0.0026. Patients gaining higher benefits from SGLT2i trended to be older, male, and overweight, compared to who were benefited less from SGLT2i.Conclusion The protective effect of SGLT2i against CKD was indicated, albeit heterogeneous. Counterfactual models including DDML-NN and conventional TE models demonstrated similar performance. However, data from a larger cohort are still required to draw a firm conclusion.
SN - 1758-5996
ER -