Journal article
Evaluation of factors predicting transition from prediabetes to diabetes among patients residing in underserved communities in the United States – A machine learning approach
Computers in biology and medicine, Vol.187, 109824
03/01/2025
DOI: 10.1016/j.compbiomed.2025.109824
PMID: 39933273
Abstract
Over one-third of the population in the United States (US) has prediabetes. Unfortunately, underserved population in the United States face a higher burden of prediabetes compared to urban areas, increasing the risk of stroke and heart disease. There is a gap in the literature in understanding early predictors of diabetes among patients with prediabetes living in underserved communities in the United States. Hence, this study's objective is to identify factors influencing the transition from prediabetes to diabetes in rural or underserved communities using a machine learning approach.
We conducted a retrospective analysis of data from prediabetic patients between 2012 and 2022. Eligible participants were at least 18 years old with baseline HbA1c levels between 5.7 % and 6.4 %. Eleven machine learning algorithms were evaluated using ten-fold cross-validation, including Logistic Regression (LR), Support Vector Classifier (SVC), K-nearest Neighbor (KNN), Gaussian Naive Bayes (GaussianNB), Bernoulli Naive Bayes (BernoulliNB), Adaptive Boosting (AdaBoost), Decision Tree (DT), Random Forest (RF), Gradient Boosting (GB), Extreme Gradient Boosting (XGBoost), and Extra Trees (ET). Subsequently, the SHAP framework was used to assess predictor influence and interactions observed with the top model.
Out of 5816 patients, 1910 met the criteria, with 426 progressing to diabetes. The Random Forest model achieved the highest accuracy (90.0 %) and AUC (0.963), followed by Extra Trees (89.5 % accuracy, AUC 0.962) and XGBoost (88.6 % accuracy, AUC 0.952). Logistic Regression demonstrated lower performance but outperformed other models such as K-Nearest Neighbors and Gaussian Naive Bayes. SHAP analysis with the RF model identified key predictors and their interactions. A significant interaction showed that lower BMI values, combined with increasing age, were associated with a reduced risk of diabetes progression, while higher BMI at younger ages increased the likelihood of progression. Additionally, several social determinants of health were identified as significant predictors.
Among the 11 models, the Random Forest model showed the strongest reliability for predicting diabetes progression. The results of this study can be used to inform public policy implications for the development of early, targeted interventions focusing on social determinants of health, dietary counseling, and BMI management to prevent diabetes in underserved communities.
•Machine learning models identified key predictors of diabetes progression.•The Random Forest model showed the best performance (90 % accuracy, AUC = 0.963).•High BMI, and lack of dietary counseling increased diabetes risk.•Race and insurance status, significantly influenced progression.•Screening and interventions are needed for high-risk groups.
Details
- Title: Subtitle
- Evaluation of factors predicting transition from prediabetes to diabetes among patients residing in underserved communities in the United States – A machine learning approach
- Creators
- Arinze Nkemdirim Okere - College of Pharmacy, The University of Iowa, 180 South Grand Ave, 366B College of Pharmacy Building (CPB), Iowa City, IA, 52242, USATianfeng Li - Florida Agricultural and Mechanical UniversityCarlos Theran - Florida Agricultural and Mechanical UniversityEunice Nyasani - WalgreensAskal Ayalew Ali - Florida Agricultural and Mechanical University
- Resource Type
- Journal article
- Publication Details
- Computers in biology and medicine, Vol.187, 109824
- Publisher
- Elsevier Ltd
- DOI
- 10.1016/j.compbiomed.2025.109824
- PMID
- 39933273
- ISSN
- 0010-4825
- eISSN
- 1879-0534
- Language
- English
- Date published
- 03/01/2025
- Academic Unit
- Pharmacy Practice and Science; Internal Medicine
- Record Identifier
- 9984845052802771
Metrics
1 Record Views