Rethinking Postpartum Glucose Assessment: Is One-Hour Testing the Key to Success? Free

Gestational diabetes mellitus (GDM) affects 14% of pregnancies worldwide (1) and is defined as any degree of glucose intolerance that is first identified during pregnancy (2). While previously unrecognized overt diabetes is ideally excluded when making the clinical diagnosis, GDM can ultimately represent prediabetes, type 2 diabetes, evolving type 1 diabetes, or even monogenic diabetes (3). Due to its superiority over HbA_1c or fasting glucose alone in detecting early-stage dysglycemia, a 2-h, 75-g oral glucose tolerance test (OGTT) is recommended for women with GDM at 4–12 weeks postpartum to ensure adequate reclassification of glucose tolerance using standard World Health Organization criteria (3). Given the elevated lifetime risk of overt diabetes following GDM (up to 60%) (4), this initial postpartum test also serves to identify people at highest risk of long-term diabetes progression. Such individuals can benefit from lifestyle interventions and, if indicated, pharmacotherapy to treat associated risk factors and delay progression to diabetes (5,6).

Clearly, the postpartum period provides a unique window of opportunity to engage in preventive health care for this high-risk population with GDM. Unfortunately, while most women return for some element of postpartum care (7), adherence to postpartum OGTT testing among those with GDM is low, with a reported average completion rate of 40% (8). Prior work has explored barriers to postpartum diabetes screening. While health care providers cite factors such as poor interdisciplinary communication and inconsistent guidelines (9), the most frequent barrier reported by women with GDM is lack of time for a 2-h test given the many competing demands of postpartum life (9,10). Individual groups have demonstrated improved adherence to postpartum glucose testing by employing strategies such as appointing a dedicated coordinator (11) or conducting the test during the delivery hospitalization (12). However, to effectively and sustainably bridge this gap in clinical care on a wider scale, a truly transformative approach is needed.

In this issue of Diabetes Care, Retnakaran et al. (13) present their analysis using data from a prospective cohort study to investigate the utility of a 1-h versus a 2-h postpartum OGTT. This study is timely, as the International Diabetes Federation recently published a position statement outlining the superiority of the 1-h (after 75-g oral glucose load) plasma glucose over other glycemic parameters (including fasting and 2-h values) in predicting diabetes and related complications (14). In fact, they provide specific 1-h glucose thresholds of ≥8.6 mmol/L and ≥11.6 mmol/L to define intermediate hyperglycemia and diabetes, respectively, and recommend that they be added to existing criteria. Retnakaran et al. (13) therefore hypothesized that a 1-h test using these glycemic thresholds might offer similar diabetes prediction in a postpartum population. In theory, this approach may be a more acceptable alternative to the current 2-h OGTT.

In this study, 369 women underwent a 2-h 75-g OGTT at 3 months, 1 year, and 5 years postpartum. At the initial 3-month postpartum test (n = 253 individuals), 70 (19.0%) individuals had prediabetes or diabetes. By 5 years postpartum, 88 (23.8%) had prediabetes or diabetes. The study population was stratified according to 1-h glucose tertile on the 3-month test. It was observed that those in the highest tertile were more likely to be of non-White ethnicity, have a higher BMI, and have experienced GDM in the preceding pregnancy. The average fasting and 2-h glucose also increased across tertiles of 1-h glucose at the 3-month time point, with a decrease in surrogate measures of whole-body insulin sensitivity (Matsuda index) and β-cell function (insulin secretion-sensitivity index-2 [ISSI-2] and insulinogenic index/HOMA of insulin resistance [IGI/HOMA-IR]). These findings remained consistent at each of the subsequent time points, indicating that stratification by tertiles of 1-h glucose was associated with metabolic risk at baseline and longitudinally.

The authors then examined the predictive capacity of the 1-h glucose at 3 months postpartum to identify current and future dysglycemia. At the 3-month OGTT, the 1-h glucose identified 60 of 70 women who were diagnosed on the 2-h glucose and identified an additional 96 women. Over time, the number of women diagnosed by 1-h glucose alone decreased, whereby at 5 years postpartum there were 48 diagnosed based on the 1-h glucose alone, 12 by the 2-h glucose alone, and 76 meeting both criteria. The top-ranked predictor of future dysglycemia was the 3-month, 1-h glucose, closely followed by the 3-month, 2-h glucose. Of note, dysglycemia was defined by the development of either World Health Organization prediabetes or diabetes criteria or the recent International Diabetes Federation1-h criteria for intermediate hyperglycemia or diabetes.

This study offers several unique strengths, including a large cohort with detailed assessment of metabolic health over 5 years postpartum. The results are supported by the strong pathophysiological changes identified by Matsuda index, ISSI-2, and IGI/HOMA-IR in the study. These data represent important progress in our understanding of glucose trajectories and diabetes evolution in postpartum women and are reassuringly consistent with studies in different populations identifying the 1-h postchallenge glucose as a predictor of future diabetes (14–16). Prior work demonstrates that individuals with an isolated 1-h glucose elevation have impaired β-cell function and an intermediate cardiovascular risk profile between those with normal 1-h glucose and impaired 2-h glucose tolerance (14–17). Therefore, 1-h glucose data could facilitate identification and treatment of high-risk individuals earlier in their diabetes evolution, before they cross standard prediabetes thresholds. Of course, in this study people with an elevated 1-h glucose frequently had other known risk factors for diabetes (more likely to be of non-White ethnicity, to have a higher BMI, and to have had previous GDM) and are therefore more likely to have significant underlying β-cell dysfunction. In general, is also important to remember that there is significant variability in OGTT testing and significant reclassification of tolerance status on retesting if glucose thresholds alone are used compared with model-based assessment of islet function. Furthermore, insulin action is difficult to measure using a 1-h OGTT (18,19).

In the current study, the inclusion of women with both normal glucose tolerance and GDM using National Diabetes Data Group criteria during the preceding pregnancy overcomes some concerns regarding the performance of this approach in settings where alternative GDM diagnostic criteria are used. On the other hand, the external validity of these findings relative to more ethnically diverse populations with a higher background prevalence of type 2 diabetes remains to be determined. It also remains to be determined if the predictive power of the 1-h glucose persists beyond 5 years of follow-up and how it correlates with long-term diabetes-related complications in postpartum populations. In addition, we do not have data demonstrating the cost versus benefit of implementing a 1-h approach, particularly on a population level, given the prevalence of isolated 1-h plasma glucose elevation is up to 16% in population-based cohort studies and 42% in high-risk cohorts (20). Finally, while moving to 1-h testing would identify more at-risk individuals at an earlier time point, a minority who would be identified on the 2-h test only would go unrecognized.

Overall, the study findings provide a theoretical groundwork for an alternative, more time-efficient approach to postpartum glucose testing. Of course, a randomized clinical trial is needed to formally test the clinical and economic outcomes associated with the implementation of a 1-h versus 2-h test. However, all such outcomes will be of minimal benefit if the 1-h test does not actually result in improved uptake among postpartum women.