Predictive utilities of lipid traits, lipoprotein subfractions and other risk factors for incident diabetes : A machine learning approach in the Diabetes Prevention Program
(2021) In BMJ Open Diabetes Research and Care 9(1).- Abstract
Introduction Although various lipid and non-lipid analytes measured by nuclear magnetic resonance (NMR) spectroscopy have been associated with type 2 diabetes, a structured comparison of the ability of NMR-derived biomarkers and standard lipids to predict individual diabetes risk has not been undertaken in larger studies nor among individuals at high risk of diabetes. Research design and methods Cumulative discriminative utilities of various groups of biomarkers including NMR lipoproteins, related non-lipid biomarkers, standard lipids, and demographic and glycemic traits were compared for short-term (3.2 years) and long-term (15 years) diabetes development in the Diabetes Prevention Program, a multiethnic, placebo-controlled, randomized... (More)
Introduction Although various lipid and non-lipid analytes measured by nuclear magnetic resonance (NMR) spectroscopy have been associated with type 2 diabetes, a structured comparison of the ability of NMR-derived biomarkers and standard lipids to predict individual diabetes risk has not been undertaken in larger studies nor among individuals at high risk of diabetes. Research design and methods Cumulative discriminative utilities of various groups of biomarkers including NMR lipoproteins, related non-lipid biomarkers, standard lipids, and demographic and glycemic traits were compared for short-term (3.2 years) and long-term (15 years) diabetes development in the Diabetes Prevention Program, a multiethnic, placebo-controlled, randomized controlled trial of individuals with pre-diabetes in the USA (N=2590). Logistic regression, Cox proportional hazards model and six different hyperparameter-tuned machine learning algorithms were compared. The Matthews Correlation Coefficient (MCC) was used as the primary measure of discriminative utility. Results Models with baseline NMR analytes and their changes did not improve the discriminative utility of simpler models including standard lipids or demographic and glycemic traits. Across all algorithms, models with baseline 2-hour glucose performed the best (max MCC=0.36). Sophisticated machine learning algorithms performed similarly to logistic regression in this study. Conclusions NMR lipoproteins and related non-lipid biomarkers were associated but did not augment discrimination of diabetes risk beyond traditional diabetes risk factors except for 2-hour glucose. Machine learning algorithms provided no meaningful improvement for discrimination compared with logistic regression, which suggests a lack of influential latent interactions among the analytes assessed in this study. Trial registration number Diabetes Prevention Program: NCT00004992; Diabetes Prevention Program Outcomes Study: NCT00038727.
(Less)
- author
- Varga, Tibor V. LU ; Liu, Jinxi ; Goldberg, Ronald B. ; Chen, Guannan ; Dagogo-Jack, Samuel ; Lorenzo, Carlos ; Mather, Kieren J. ; Pi-Sunyer, Xavier ; Brunak, Søren and Temprosa, Marinella
- organization
- publishing date
- 2021
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- diabetes mellitus, lipids, lipoproteins, prediabetic state, type 2
- in
- BMJ Open Diabetes Research and Care
- volume
- 9
- issue
- 1
- article number
- 001953
- publisher
- BMJ Publishing Group
- external identifiers
-
- pmid:33789908
- scopus:85103629343
- ISSN
- 2052-4897
- DOI
- 10.1136/bmjdrc-2020-001953
- language
- English
- LU publication?
- yes
- id
- e0a15ee6-cce5-48d6-a0da-407f9346674a
- date added to LUP
- 2021-04-12 13:03:32
- date last changed
- 2025-02-09 09:51:10
@article{e0a15ee6-cce5-48d6-a0da-407f9346674a, abstract = {{<p>Introduction Although various lipid and non-lipid analytes measured by nuclear magnetic resonance (NMR) spectroscopy have been associated with type 2 diabetes, a structured comparison of the ability of NMR-derived biomarkers and standard lipids to predict individual diabetes risk has not been undertaken in larger studies nor among individuals at high risk of diabetes. Research design and methods Cumulative discriminative utilities of various groups of biomarkers including NMR lipoproteins, related non-lipid biomarkers, standard lipids, and demographic and glycemic traits were compared for short-term (3.2 years) and long-term (15 years) diabetes development in the Diabetes Prevention Program, a multiethnic, placebo-controlled, randomized controlled trial of individuals with pre-diabetes in the USA (N=2590). Logistic regression, Cox proportional hazards model and six different hyperparameter-tuned machine learning algorithms were compared. The Matthews Correlation Coefficient (MCC) was used as the primary measure of discriminative utility. Results Models with baseline NMR analytes and their changes did not improve the discriminative utility of simpler models including standard lipids or demographic and glycemic traits. Across all algorithms, models with baseline 2-hour glucose performed the best (max MCC=0.36). Sophisticated machine learning algorithms performed similarly to logistic regression in this study. Conclusions NMR lipoproteins and related non-lipid biomarkers were associated but did not augment discrimination of diabetes risk beyond traditional diabetes risk factors except for 2-hour glucose. Machine learning algorithms provided no meaningful improvement for discrimination compared with logistic regression, which suggests a lack of influential latent interactions among the analytes assessed in this study. Trial registration number Diabetes Prevention Program: NCT00004992; Diabetes Prevention Program Outcomes Study: NCT00038727. </p>}}, author = {{Varga, Tibor V. and Liu, Jinxi and Goldberg, Ronald B. and Chen, Guannan and Dagogo-Jack, Samuel and Lorenzo, Carlos and Mather, Kieren J. and Pi-Sunyer, Xavier and Brunak, Søren and Temprosa, Marinella}}, issn = {{2052-4897}}, keywords = {{diabetes mellitus; lipids; lipoproteins; prediabetic state; type 2}}, language = {{eng}}, number = {{1}}, publisher = {{BMJ Publishing Group}}, series = {{BMJ Open Diabetes Research and Care}}, title = {{Predictive utilities of lipid traits, lipoprotein subfractions and other risk factors for incident diabetes : A machine learning approach in the Diabetes Prevention Program}}, url = {{http://dx.doi.org/10.1136/bmjdrc-2020-001953}}, doi = {{10.1136/bmjdrc-2020-001953}}, volume = {{9}}, year = {{2021}}, }