In July 2025, the journal Diabetes Care published the results of an observational longitudinal analysis evaluating the ability of a 1-hour oral glucose tolerance test (OGTT) to predict the risk for progression to prediabetes when given to people with normal glucose tolerance. This study included nearly 1,600 participants from the San Antonio Heart Study without diabetes at baseline, who had an initial OGTT and a subsequent OGTT 7.5 years later.
The single strongest predictor for developing prediabetes was the 1-hour OGTT. Overall, 24.7% of participants had progressed to prediabetes; 22.5% with a 1-hour OGTT glucose of < 155 mg/dL, but 42.5% of those with a glucose of > 155 mg/dL. Using 120 mg/dL as a cut-point yielded a 61% sensitivity and 67% specificity for identifying people at high risk for developing prediabetes. This provides a reasonable target to identify people at risk for prediabetes quite early, at a stage highly likely to respond to preventative measures. It’s also important to note that most clinicians use the 2-hour OGTT, and this study points to the superior predictive value of the 1-hr OGTT for determining prediabetes risk. This is not the first study to do so; the International Diabetes Federation’s position statement advocates for the 1-hour OGTT, considering it superior for identifying risk for diabetes and other chronic conditions, and diagnostic for type 2 diabetes at glucose levels of 209 mg/dL or higher.
It’s worth noting that despite their convenience, a fasting glucose or HbA1c alone can miss a substantial amount of risk. For example, in the NIH’s 3rd edition of Diabetes in America, in terms of type 2 diabetes diagnosis, an OGTT is clearly superior to both. Using NHANES data, an HbA1c test detects approximately 30% of the people who have undiagnosed diabetes and a fasting glucose test identifies 47%, while the 2-hour OGTT catches over 90% of the undiagnosed. From another angle, 47% of undiagnosed diabetes is only detected by an OGTT and would be missed with a fasting glucose and HbA1c. Considering ~25 to 35% of all cases of type 2 diabetes are undiagnosed, reliance on only HbA1c or fasting glucose would miss nearly half of them. And the prevalence of undiagnosed diabetes is even higher in some age groups; just over 40% of all diabetes remains undiagnosed in 20-44 year olds, for example, and reaches a similar percentage in adults over 75, the second highest at-risk age group for undiagnosed diabetes.
Despite the clear advantages of the OGTT, a second recently published study points to the value of fasting plasma glucose for identifying prediabetes in some populations. Published in BMJ Open Diabetes Research & Care, a cohort of nearly 600 participants with a family history of young-onset type 2 diabetes (YOD, defined as a diagnosis before age 40) was followed for just over 12 years. Having a positive family history was associated with a nearly 5-fold risk for developing diabetes in this study.
Among participants without a family history of YOD, a 1-hour OGTT remained a strong predictor for the risk of developing diabetes. However, fasting plasma glucose outperformed the 1-hour OGTT among those with a family history of YOD. The Area Under the Receiver Operating Characteristic Curve, or AUROC, is a metric that plots the true positive rate against the false positive rate, with 1 being the perfect score, and 0.5 equivalent to random guessing. The AUROC for fasting plasma glucose was 0.792, higher than the 1-hour OGTT (0.688) as well as the 2-hour OGTT (0.706). The predictive value of the 1-hour OGTT was mostly attenuated in this population, i.e., among people with a family history of YOD, a fasting plasma glucose may still be the best test for identifying risk.
The authors of this study point out that the difference in the predictive value of these tests in people with and without a family history of YOD indicates a potential difference in pathophysiology. For example, it is more likely that impaired fasting glucose reflects impaired hepatic insulin sensitivity or beta-cell dysfunction or low beta-cell mass, whereas impaired glucose tolerance reflects reduced peripheral insulin sensitivity and a progressive loss of beta-cell function. We do have evidence that a 1-hour OGTT is superior to the 2-hour OGTT for detecting deterioration of beta-cell function, perhaps explaining its better predictive value among people with impaired glucose tolerance. There is also some evidence that the pathophysiology behind impaired fasting glucose responds less well to lifestyle modifications than impaired glucose tolerance; thus, these different diagnostic methods not only differ in their risk prediction value but may also provide guidance for optimal therapies.
Lastly, another diagnostic test deserves mention: the triglyceride glucose (TyG) index. Easily calculated from fasting triglyceride and glucose levels, it is thought to reflect insulin resistance, which underlies many metabolic disorders. An umbrella review published in Cardiovascular Diabetology found an elevated TyG was associated with kidney disease, risk of contrast-induced nephropathy, type 2 and gestational diabetes, diabetic retinopathy, metabolic syndrome, obstructive sleep apnea, cognitive impairment, and more. A strong and early predictor of metabolic risk, this test is an easy calculation from labs and already likely available.