Call for Modification of the Intermediate Hyperglycemic classification of the 1hour-Oral Glucose Tolerance Test: Insight from the Africans in America cohort Commentary

Emerging evidence from the International Diabetes Federation (IDF) Position Statement of 2024 suggests the 1-hour oral glucose tolerance test (1-h OGTT) provides earlier detection of diabetes, diabetes risk and cardiometabolic disease than the 2-h OGTT(1).

Yet 1-h OGTT and 2-h OGTT share an important characteristic. Diagnostically they each have three categories(1, 2). However, the former relies solely on the 1-h glucose values, while the latter uses both fasting plasma glucose (FPG) and 2-h glucose values(1, 2). The names of the three categories for the 1-h OGTT are: normal glucose tolerant (NGT) (1-h glucose<155 mg/dL), intermediate hyperglycemia (IH) (1-h glucose≥155 & <209 mg/dL) and diabetes (I-h glucose≥209 mg/dL)(1). The three 2-h OGTT categories are: NGT (FPG<100 mg/dL & 2-h OGTT<140 mg/dL), prediabetes (FPG≥100 & <126 mg/dl and/or 2-h glucose≥140 mg/dL & <200 mg/dL) and diabetes (FPG≥126 mg/dL and/or 2-h glucose≥200 mg/dL)(2). Both the IH category in the 1-h OGTT and the prediabetes category in the 2-h OGTT describe the degree of glycemia which occurs between their respective definitions of NGT and diabetes and has recently been referred to as preDM/IH(3).

However, a methodologic difference between IH and prediabetes exists. IH stands as a single group, whereas, prediabetes is divided into three groups: impaired fasting glucose (IFG) (FPG≥100 mg/dL), impaired glucose tolerance (IGT) (2-h glucose≥140 mg/dL & <200 mg/dL), and IFG and IGT combined. As cardiometabolic risk increases across the prediabetes groups and the prediabetes classification leverages information from FPG, we hypothesized that risk detection in the 1-h OGTT might be enhanced by sub-dividing the IH group based on FPG status(4).

We raise this possibility because in a previous analysis of African immigrants in the United States, 33% of the individuals with diabetes based on 2-h glucose≥200 mg/dL were not detected by the 1-h OGTT of ≥209 mg/dL(5). By 1-h OGTT criteria they were classified as IH(5). We suspect that individuals diagnosed with diabetes by a 2-h OGTT, but IH by a 1-h OGTT, are at higher cardiometabolic risk than individuals in the IH group who did not meet the 2-h threshold of ≥200 mg/dL.

To test this hypothesis, we conducted a pilot study using existing 1-h OGTT data from 645 enrollees in the Africans in America cohort (male 62%, age 39±11 (mean±SD), range 20–70y, BMI 27.8±4.6, range 18.8–46.2kg/m²) and subdivided the IH category into groups according to the presence of IFG (≥100 mg/dL & <126 mg/dL)(6–8). In short, four groups were created: NGT; IH+normal fasting glucose (IH+NFG); IH+impaired fasting glucose (IH+IFG); and diabetes (Figure, Supplement Table).

Figure: Cardiometabolic Characteristics by 1-h OGTT Sub-Groups.

Panels A & B: Insulin Sensitivity measured by Matsuda Index; C & D: Atherosclerotic Cardiovascular Disease Risk Score (ASCVD)≥7.5%; E & F: Metabolic Dysfunction-Associated Steatotic Liver Disease (MASLD) prevalence. Data presented in Panels A & B are mean±SE bars and in Panels C through F as percent.

For continuous variables both unpaired t-tests and Mann-Whitney tests were used to compare cardiometabolic risk factors in the IH+NFG and IH+IFG groups. Chi-square tests were used for categorical variables. The Jonckheere–Terpstra test was used to determine the trend of each cardiometabolic variable across all of the glycemic 1-h OGTT groups (Figure, Supplement Table).

Variables were divided into four categories (Supplement Table): (1) demographics (sex, age, BMI, blood pressure, waist circumference and visceral adipose tissue (VAT) measured by computerized tomographic scan(9); (2) physiologic characteristics (fasting glucose, area under the glucose curve and insulin-sensitivity as determined by the Matsuda Index; (3) cardiovascular (lipids, LDL particle number, atherosclerotic cardiovascular disease (ASCVD) risk(10), and (4) metabolic (percent with metabolic-dysfunction associated steatotic liver disease (MASLD)(11, 12).

The glucose tolerance status of the 645 enrollees were: NGT 58% (375/645); IH+NFG 27% (175/645); IH+IFG 7% (41/645), and diabetes 8% (54/645). Compared to the IH-NGT, the IH-IFG were older, had higher BMI, greater glycemia and less insulin-sensitivity (i.e lower Matsuda Index values), lower HDL, higher LDL particle number, higher ASCVD risk score, and a higher prevalence of MASLD (all P<0.05) (Figure, Supplement Table). Furthermore, when examined across the 4 groups, every variable increased significantly (all P≤0.01).

The concept that cardiometabolic risk is higher in the IH-IFG group than the IH-NFG group is supported by that observation that insulin-sensitivity is lower and both ASCVD risk and MASLD prevalence are higher in the IH-IFG group (all P<0.001) (Figure). VAT volume and LDL particle number, both of which are linked to decreased insulin-sensitivity, were higher in the IH-IFG than the IH-NFG(13, 14) (Supplement Table). All cardiometabolic variables trended significantly upward across the four glucose tolerance groups (Supplement Table).

As this proof-of-concept analysis used previously collected cross-sectional data in a single population, it cannot by itself, inform policy. However, we suggest that the pattern of the results presented from this pilot study of Africans in America data is sufficiently compelling to justify the investment of resources to re-examine across populations the IH data from the 15 studies in the original meta-analysis which the IDF used to identify the 1-h glucose threshold for diagnosis of diabetes(15). Data from the studies (4 longitudinal and 11 cross-sectional) are already curated and available, and report age and BMI. Furthermore, working with the data from the longitudinal studies, the prognostic value of the two IH groups could be ascertained. Additional data available from the 15 studies include insulin resistance, lipids, blood pressure and cardiovascular risk. While none of the studies provide information on MASLD, data relating MASLD to 1-h OGTT parameters are available(16, 17).

As Ahuja et al. recognized that the meta-analyses they conducted had no studies from South America or Africa, this IH evaluation could be expanded to include cohorts from South American and African cohorts(15). Data recently obtained from Brazil are now available(3). Additionally, previously published data from the Africans in America cohort could be accessed(5, 18). We are not aware of other African cohorts which have data on 1-h glucose values during an OGTT.

Overall, our proof-of-concept pilot study with Africans in America data has demonstrated that subdividing the IH group by the presence of IFG distinguishes cardiometabolic risk and may justify increased intervention for the subset with IH+IFG. As measurement of FPG is already part of the 1-h OGTT protocol, detection of risk may be enhanced without increasing cost or burden. If existing data analyses demonstrate the value across populations of subdividing the IH by fasting glucose levels, treatment paradigms and policy could be optimized.

Data Sharing Statement

De-identified participant data will be made available in Excel format upon reasonable request to the corresponding author, Anne E Sumner, MD.