Machine learning approaches revealed metabolic signatures of incident chronic kidney disease in persons with pre-and type 2 diabetes

Early and precise identification of individuals with pre-diabetes and type 2 diabetes (T2D) at risk of progressing to chronic kidney disease (CKD) is essential to prevent complications of diabetes. Here, we identify and evaluate prospective metabolite biomarkers and the best set of predictors of CKD in the longitudinal, population-based Cooperative Health Research in the Region of Augsburg (KORA) cohort by targeted metabolomics and machine learning approaches. Out of 125 targeted metabolites, sphingomyelin (SM) C18:1 and phosphatidylcholine diacyl (PC aa) C38:0 were identified as candidate metabolite biomarkers of incident CKD specifically in hyperglycemic individuals followed during 6.5 years. Sets of predictors for incident CKD developed from 125 metabolites and 14 clinical variables showed highly stable performances in all three machine learning approaches and outperformed the currently established clinical algorithm for CKD. The two metabolites in combination with five clinical variables were identified as the best set of predictors and their predictive performance yielded a mean area value under the receiver operating characteristic curve of 0.857. The inclusion of metabolite variables in the clinical prediction of future CKD may thus improve the risk prediction in persons with pre- and T2D. The metabolite link with hyperglycemia-related early kidney dysfunction warrants further investigation.

Download Full-text

Machine learning approaches revealed metabolic signatures of incident chronic kidney disease in persons with pre-and type 2 diabetes

10.2337/figshare.13022624.v1 ◽

2020 ◽

Author(s):

Ada Admin ◽

Jialing Huang ◽

Cornelia Huth ◽

Marcela Covic ◽

Martina Troll ◽

...

Keyword(s):

Machine Learning ◽

Type 2 Diabetes ◽

Chronic Kidney Disease ◽

Kidney Disease ◽

Characteristic Curve ◽

Predictive Performance ◽

Population Based ◽

Learning Approaches ◽

Clinical Variables

Early and precise identification of individuals with pre-diabetes and type 2 diabetes (T2D) at risk of progressing to chronic kidney disease (CKD) is essential to prevent complications of diabetes. Here, we identify and evaluate prospective metabolite biomarkers and the best set of predictors of CKD in the longitudinal, population-based Cooperative Health Research in the Region of Augsburg (KORA) cohort by targeted metabolomics and machine learning approaches. Out of 125 targeted metabolites, sphingomyelin (SM) C18:1 and phosphatidylcholine diacyl (PC aa) C38:0 were identified as candidate metabolite biomarkers of incident CKD specifically in hyperglycemic individuals followed during 6.5 years. Sets of predictors for incident CKD developed from 125 metabolites and 14 clinical variables showed highly stable performances in all three machine learning approaches and outperformed the currently established clinical algorithm for CKD. The two metabolites in combination with five clinical variables were identified as the best set of predictors and their predictive performance yielded a mean area value under the receiver operating characteristic curve of 0.857. The inclusion of metabolite variables in the clinical prediction of future CKD may thus improve the risk prediction in persons with pre- and T2D. The metabolite link with hyperglycemia-related early kidney dysfunction warrants further investigation.

Download Full-text

Machine Learning Approaches Reveal Metabolic Signatures of Incident Chronic Kidney Disease in Individuals With Prediabetes and Type 2 Diabetes

Diabetes ◽

10.2337/db20-0586 ◽

2020 ◽

Vol 69 (12) ◽

pp. 2756-2765

Author(s):

Jialing Huang ◽

Cornelia Huth ◽

Marcela Covic ◽

Martina Troll ◽

Jonathan Adam ◽

...

Keyword(s):

Machine Learning ◽

Type 2 Diabetes ◽

Chronic Kidney Disease ◽

Kidney Disease ◽

Learning Approaches ◽

Incident Chronic Kidney Disease ◽

Metabolic Signatures

Download Full-text

Diagnosis and Prediction of Type-2 Chronic Kidney Disease Using Machine Learning Approaches

10.1201/9781003185604-5 ◽

2021 ◽

pp. 79-90

Author(s):

Ritu Aggarwal ◽

Prateek Thakral

Keyword(s):

Machine Learning ◽

Chronic Kidney Disease ◽

Kidney Disease ◽

Learning Approaches

Download Full-text

Glycemic Control and the Risk of Acute Kidney Injury in Patients With Type 2 Diabetes and Chronic Kidney Disease: Parallel Population-Based Cohort Studies in U.S. and Swedish Routine Care

Diabetes Care ◽

10.2337/dc20-1588 ◽

2020 ◽

Vol 43 (12) ◽

pp. 2975-2982

Author(s):

Yang Xu ◽

Aditya Surapaneni ◽

Jim Alkas ◽

Marie Evans ◽

Jung-Im Shin ◽

...

Keyword(s):

Type 2 Diabetes ◽

Chronic Kidney Disease ◽

Acute Kidney Injury ◽

Kidney Disease ◽

Glycemic Control ◽

Cohort Studies ◽

Kidney Injury ◽

Population Based ◽

Routine Care

Download Full-text

High prevalence of chronic kidney disease in population-based patients diagnosed with type 2 diabetes in downtown Shanghai

Journal of Diabetes and its Complications ◽

10.1016/j.jdiacomp.2007.08.001 ◽

2008 ◽

Vol 22 (2) ◽

pp. 96-103 ◽

Cited By ~ 34

Author(s):

Bin Lu ◽

Xiaoyan Song ◽

Xuehong Dong ◽

Yehong Yang ◽

Zhaoyun Zhang ◽

...

Keyword(s):

Type 2 Diabetes ◽

Chronic Kidney Disease ◽

Kidney Disease ◽

Population Based ◽

High Prevalence

Download Full-text

Multiple risk factor control, mortality and cardiovascular events in type 2 diabetes and chronic kidney disease: a population-based cohort study

BMJ Open ◽

10.1136/bmjopen-2017-019950 ◽

2018 ◽

Vol 8 (5) ◽

pp. e019950 ◽

Cited By ~ 10

Author(s):

Shota Hamada ◽

Martin C Gulliford

Keyword(s):

Type 2 Diabetes ◽

Chronic Kidney Disease ◽

Risk Factor ◽

Cohort Study ◽

Kidney Disease ◽

Cardiovascular Events ◽

Population Based ◽

Multiple Risk Factor ◽

Risk Factor Control

Download Full-text

Characteristics, Mortality and Cardiovascular Events in Chronic Kidney Disease According To Previous Type 2 Diabetes Mellitus and/or Hypertension. A Population-Based Epidemiological Study (KIDNEES)

10.21203/rs.3.rs-658979/v1 ◽

2021 ◽

Author(s):

Betlem Salvador-González ◽

Oriol Cunillera-Puértolas ◽

David Vizcaya ◽

M Jesus Cerain-Herrero ◽

Neus Gil-Terrón ◽

...

Keyword(s):

Diabetes Mellitus ◽

Type 2 Diabetes ◽

Chronic Kidney Disease ◽

Type 2 Diabetes Mellitus ◽

Kidney Disease ◽

Cardiovascular Events ◽

Population Based ◽

Adverse Outcomes

Abstract Introduction and objectives. Chronic Kidney Disease (CKD) entails a considerable burden of adverse outcomes. Identifying the cause is recommended but data on its prognostic value are scarce. We aimed to estimate how the clinical, cardiovascular events (CVE) and all-cause mortality (ACM) of CKD patients differs according to previous Type 2 Diabetes Mellitus (2TD) and/or Hypertension (HTN). Methods. We conducted a retrospective cohort study based on electronic health records of subjects aged 18–90 years old, with incident CKD between 1st January 2007 and 31st December 2017. The association between CKD groups according to previous T2D and/or HTN, and risk of ACM and CVE at follow-up were determined with Cox and Fine-Gray regressions, respectively. Results. 398,477 subjects were included. Median age was 74years, 55.2% were women. Individuals were distributed to HTN-CKD (51.9%), T2D-CKD (3.87%), HTN/T2D-CKD (31.4%) and unspecified-CKD (12.9%). In the multivariate analysis, with the T2D-CKD group as reference, the ACM Hazard Ratio (HR) was 0.645 (95%CI 0.624–0.667) in HTN-CKD, 0.704 (95%CI 0.682–0.728) in HTN/T2D-CKD and 0.875 (95%CI 0.844–0.908) in Unspecified-CKD group. The respective sub distribution HRs for CVE were 1.006 (CI95% 0.946–1.069), 1.238 (CI95% 1.164–1.316) and 0.722 (CI95% 0.665–0.785). Conclusion. In individuals with CKD, the risk of ACM and CVE differed according to previous HTN or/and T2D. These characteristics can help identifying individuals at higher risk of adverse outcomes, and improving the management of CKD patients in primary care.

Download Full-text