Diabetes Mellitus Risk Prediction in the Framingham Offspring Study and Large Population Analysis
Issued Date
2025-04-01
Resource Type
eISSN
20726643
Scopus ID
2-s2.0-105002383568
Journal Title
Nutrients
Volume
17
Issue
7
Rights Holder(s)
SCOPUS
Bibliographic Citation
Nutrients Vol.17 No.7 (2025)
Suggested Citation
Ai M., Otokozawa S., Liu C.T., Asztalos B.F., Maddalena J., Diffenderfer M.R., Russo G., Thongtang N., Dansinger M.L. Diabetes Mellitus Risk Prediction in the Framingham Offspring Study and Large Population Analysis. Nutrients Vol.17 No.7 (2025). doi:10.3390/nu17071117 Retrieved from: https://repository.li.mahidol.ac.th/handle/123456789/109621
Title
Diabetes Mellitus Risk Prediction in the Framingham Offspring Study and Large Population Analysis
Author's Affiliation
Institute of Science Tokyo
Siriraj Hospital
Sapporo Medical University School of Medicine
Jean Mayer USDA Human Nutrition Research Center on Aging
Framingham Heart Study
School of Public Health
Università degli Studi di Messina, Facoltà di Medicina e Chirurgia
Perennial Climate Inc.
Boston Heart Diagnostics
Siriraj Hospital
Sapporo Medical University School of Medicine
Jean Mayer USDA Human Nutrition Research Center on Aging
Framingham Heart Study
School of Public Health
Università degli Studi di Messina, Facoltà di Medicina e Chirurgia
Perennial Climate Inc.
Boston Heart Diagnostics
Corresponding Author(s)
Other Contributor(s)
Abstract
Background: Diabetes mellitus is a major cause of death and a significant risk factor for cardiovascular disease, kidney failure, neuropathy, and retinopathy. Our objectives were to develop a diabetes risk model and apply it to a large population. Methods: Non-diabetic adults in the Framingham Offspring Study (n = 2416) were followed for 10 years for new diabetes. At baseline, the fasting serum glucose, adiponectin, insulin, glycated albumin, total cholesterol, triglycerides (TG), and high-density lipoprotein cholesterol (HDL-C) were measured using standardized automated assays. Standard health information was collected. Diabetes risk prediction models were developed using logistic regression analysis and applied to a large population (n = 133,764). Results: In this prospective study, 166 subjects (6.9%) developed new-onset diabetes. Glucose, body mass index (BMI), log adiponectin, % log glycated albumin, parental diabetes, TG, and the use of cholesterol-lowering medications entered the model (C statistic: 0.924; 0.898, biochemical variables: 0.898, and fasting glucose: only 0.876). In the population in non-diabetic subjects (56.3) and prediabetic subjects (36.2%), the predicted 10-year diabetes risk rates were 0.4% and 5.5% with the biochemical model, respectively. Prediabetic and diabetic subjects were insulin-resistant compared to non-diabetic subjects, but only those with diabetes had significant reductions in their insulin production. Conclusions: The 10-year risk of diabetes can be accurately predicted and applied to large populations. Fasting glucose alone is diagnostic for diabetes and is an excellent predictor of future diabetes, with having prediabetes increasing the risk 6-fold. Insulin and C-peptide measurements are useful in diabetic subjects to detect decreased insulin production and the need for insulin therapy.