Detection and Intervention: Use of Continuous Glucose Monitoring in the Early Stages of Type 2 Diabetes

Dysglycemia is a significant risk factor for progressing to clinical type 2 diabetes. Approximately 5–10% of adults with moderate-to-advanced dysglycemia progress to type 2 diabetes annually (26). An estimated 25–50% of individuals with an A1C of 6.0–6.5% (42–48 mmol/mol) will progress to clinical type 2 diabetes within 5 years (27). Ligthart et al. (28) reported that 74% of individuals 45 years of age with a diagnosis of prediabetes will progress to clinical type 2 diabetes in their lifetime. Yet, as discussed above, the large majority of these people remain unaware and mostly untreated.

Effectively addressing the growing type 2 diabetes epidemic requires that we elevate dysglycemia to a higher level of priority when it is detected in combination with other known risk factors for diabetes (e.g., overweight/obesity, family history of type 2 diabetes, and sedentary lifestyle). In other words, the medical community needs to start calling it what it is: type 2 diabetes in its early stages. Some experts have suggested a change in prediabetes nomenclature to bring focus to those at higher risk of progression.

This article proposes three stages of dysglycemia, emphasizing the greater risk of individuals in stage 2 for progression to type 2 diabetes and its associated complications: stage 1 for prediabetes, stage 2 for early diabetes (preclinical diabetes), and stage 3 for clinical type 2 diabetes (Table 1). According to recent American Diabetes Association (ADA) clinical practice recommendations, individuals with prediabetes (stage 1) would require initial treatment with lifestyle interventions (i.e., weight loss and physical activity), whereas those with early diabetes (stage 2) should receive more intensive interventions (i.e., lifestyle changes in combination with medication therapy to prevent or slow progression) (29).

There is ample evidence to support interventions to reduce dysglycemia and subsequent tissue damage in individuals with stage 1 or stage 2 dysglycemia. For example, in a data-pooling analysis of nine studies from five countries with 44,623 participants, Colagiuri et al. (30) investigated glycemic thresholds for diabetes-specific retinopathy. Investigators observed a low retinopathy prevalence with a fasting plasma glucose level <108 mg/dL (<6.0 mmol/L) and an A1C <6.0% (<42 mmol/mol) but increased prevalence above those thresholds. An earlier analysis of the U.K. Prospective Diabetes Study (31) showed that incidence rates for myocardial infarction, microvascular complications, and any diabetes-related event increased in a linear manner at A1C levels ≥5.5% (≥37 mmol/mol).

Moreover, there is precedence for staging type 2 diabetes. In 2015, JDRF, the Endocrine Society, and the ADA introduced a new approach for staging of type 1 diabetes to help investigators optimize the design of their type 1 diabetes prevention trials: stage 1, ≥2 islet autoantibodies with normal glycemia; stage 2, dysglycemia; and stage 3, clinical type 1 diabetes (32). Other examples include current staging of hypertension and cancer for both diagnosis and treatment.

Formalizing our definition of type 2 diabetes to include all levels of dysglycemia, with accompanying risk factors, would help to create a greater sense of urgency among clinicians for initiating more widespread screening and more aggressive interventions when dysglycemia is detected. This, in turn, may prompt greater adherence to prescribed interventions when individuals understand the seriousness of their condition. Moreover, support for this approach within the medical community could prompt both public and commercial payers to expand their coverage eligibility for medications and technology devices that are currently made available only to individuals with a clinical diagnosis of type 2 diabetes.

The three commonly used methods of detecting dysglycemia are measurement of A1C, fasting plasma glucose (FPG), and 2-hour postprandial glucose (PG) during a 75-g oral glucose tolerance test (OGTT). However, each method has certain limitations.

Because A1C testing provides only an average glucose value over a 2- to 3-month period, it does not reflect inter- and intraday glucose excursions, which are stronger markers for dysglycemia. Moreover, the accuracy of A1C test results can be affected by numerous interferents and conditions, including anemia, sickle cell disease, age, pregnancy, race/ethnicity, uremia, dyslipidemia, and many others (33).

Although the 2-hour, 75-g OGTT is usually considered the most accurate method for detecting dysglycemia, it is a cumbersome test that requires individuals to fast overnight, provide a fasting blood sample, ingest the oral glucose load over 5 minutes, remain at a low activity level for 2 hours, and provide blood samples 1 and 2 hours post-load. The OGTT also requires confirmatory testing because of its high rate of intraindividual variability. The test results can be significantly affected by an acute episode of stress, illness, or exercise (34). It is both unpleasant and inconvenient for most individuals, particularly children.

Use of the FPG test is more convenient and less costly than the other methods. However, a major disadvantage is that a single test cannot confirm a diagnosis of dysglycemia; individuals must return to their clinician’s office on a different day to repeat the test. Moreover, as reported by Aekplakorn et al. (35), the specificity of FPG testing for detecting dysglycemia is relatively low, ranging from 47.7% to only 62.0% depending on age, sex, BMI, and family history of type 2 diabetes. One reason for this low specificity may be that elevated FPG is often detected late in the progression of dysglycemia, which makes it more difficult to reverse.

Numerous studies have demonstrated the accuracy of CGM in detecting early dysglycemia compared with other short-term methods discussed above (36). In an early evaluation, Chen et al. (37) looked at the accuracy of CGM during an OGTT in 49 outpatients with an FPG level of 70–198 mg/dL (3.9–11.0 mmol/L). The investigators observed strong correlations between CGM and venous blood glucose values throughout the OGTT tests (0.928), as well as during phases of stable (0.901), rapidly rising (0.924), and rapidly falling (0.902) glucose (P <0.001 for all). Madhu et al. (38) used CGM data to assess 60 people with obesity who were first-degree relatives of individuals with type 2 diabetes. Twenty participants had normal glucose tolerance (NGT). The investigators reported that three (15%) of the NGT participants experienced glucose excursions in the diabetes range, and 18 (90%) experienced excursions into the impaired glucose tolerance range, with the maximum excursion of 176 mg/dL (9.8 mmol/L).

Although CGM metrics for detecting dysglycemia have yet to be established, we can take some guidance from Chan et al. (39), who investigated the relationship between CGM values, A1C, and OGTT results in 98 youth with obesity and early dysglycemia (Table 2).

The ADA clinical practice guidelines recommend screening for individuals with any of the following characteristics: overweight or obesity (BMI 25–30 and >30 kg/m², respectively), age ≥35 years, sedentary lifestyle, immediate family member with type 2 diabetes, history of cardiovascular disease, hypertension, abnormal HDL cholesterol and triglyceride levels, polycystic ovary syndrome, prior gestational diabetes, HIV, or high-risk ethnicity (40). Prediabetes or early diabetes (stage 1 or stage 2) is detected when values of one or more of the three tests of dysglycemia presented in Table 1 fall into the respective diagnostic ranges. Although clinicians may want to use one of these traditional methods for initial screening, using standardized CGM data reports for confirmatory testing can help individuals better comprehend their glycemic status, which in turn creates opportunities for more productive discussion about the importance of halting or slowing progression of stage 1 or stage 2 dysglycemia to clinical type 2 diabetes (stage 3).

Before a visit with a patient suspected of being at risk, ask staff to obtain a record of the patient’s most recent A1C. If the A1C was not measured within 3 months of the appointment, staff can ask patients to go to a local laboratory in advance of the visit to get their A1C measured, along with their urine albumin-to-creatinine ratio (UACR), estimated glomerular filtration rate, thyroid-stimulating hormone, lipids levels, and a comprehensive metabolic panel that includes liver function tests. If these laboratory values are up-to-date in the patient’s chart, it may be more convenient for the patient to come to the clinic for a point-of-care A1C test, if available, before the scheduled visit.

At the scheduled visit, it is important to obtain and document a comprehensive medical history to help determine the patient’s diabetes risk. Be sure to ask about a family history of diabetes. With female patients, it is also important to ask about prior gestational diabetes or the birth weights of any children. Document all comorbid conditions, including hypertension, hyperlipidemia, chronic kidney disease, history of stroke or myocardial infarction, neuropathy (i.e., numbness or tingling), and retinopathy (i.e., vision changes) detected on a previous eye exam. This information will provide a good starting point for determining the person’s diabetes risk and early staging.

If a deeper diagnostic dive is needed, several options are available. Although the 2-hour, 75-g OGTT will be the most accurate determination of dysglycemia, as previously mentioned, this test is both time-consuming and resource-consuming, and it may not be agreeable to or feasible for many individuals. This is where CGM can come into play.

CGM devices sample interstitial glucose levels every 1–5 minutes and record and display numerical and graphical information about a person’s current glucose level and glucose trends, including trend arrows that indicate the direction and velocity of changing glucose levels. Clinicians have two options for initiating CGM: professional and personal CGM systems.

Professional CGM systems are owned by the clinic and intended for intermittent use by patients for up to 2 weeks. Use of these systems allows clinicians to analyze CGM data retrospectively to detect and pinpoint early dysglycemia (fasting, postprandial, or both). The CGM sensor is inserted in the clinic, and the handheld data reader is left at the clinic so patients cannot access their data until they return for their follow-up visit. If patients are using the Dexcom professional CGM system in the unblinded mode, they will download the Dexcom app on their smartphone. No patient training is needed. Patients using the Dexcom professional CGM for the unblinded mode will need to install an app for their phone.

Most professional devices are blinded, but one can be used in either blinded or unblinded mode. Blinded use allows for a more realistic view of glycemic status during a “regular day” and is more representative of a patient’s typical patterns of diet, exercise, and medication use with minimal influence of the wearer (41). However, we prefer using professional CGM in the unblinded mode, which allows patients to engage with the data and begin to understand the relationship between their various lifestyle behaviors (e.g., eating and exercise) and their glucose levels throughout the day and often begin making changes on their own (42). However, it will be important to carefully assess during the first few days of wearing a CGM sensor to evaluate the initial level of glycemic control before the patient may have made behavioral changes based on the CGM data.

Another option is personal CGM, which is equivalent to professional CGM when used in the unblinded mode and allows patients to view their glucose in real time. Unlike when using professional CGM, patients are encouraged to frequently view and engage with their data when using personal CGM. One advantage of personal CGM systems is that they feature programmable alarms and alerts. Another advantage with newer systems is that they do not require the purchase of a reader or transmitter and instead can be used with a smartphone app, thus lowering the startup cost.

Depending on the patient, clinicians may want to consider setting a high glucose alert at 140 mg/dL, which is the lower threshold for stage 1 dysglycemia, simply to make the patient aware that there is something going on. Patients should be instructed to pay attention to any behavior (e.g., eating or physical activity) that may be causing glucose to rise but to not try to treat it. The initial goal is to establish a baseline of daily glucose patterns to identify times and behaviors that need to be addressed. Patients should receive basic instruction on journaling the effects of food, exercise, and stress on their glucose levels. A follow-up visit to download and review the CGM data should be scheduled 2–3 weeks after initiating personal CGM.

At this time, insurance plans do not cover CGM for prediabetes. Thus, individuals will need to pay out-of-pocket for sensors. However, the U.S. Food and Drug Administration is currently evaluating a prediabetes approval application.

All CGM systems have corresponding downloading software that features various representations of the ambulatory glucose profile (AGP) report, which presents glycemic data in a graphical format. This report provides information about time in range (%TIR; the percentage of time a person spends within the target glycemic range of 70–180 mg/dL [3.9–10.0 mmol/L], time below range (%TBR; the percentage of time with glucose levels 54–69 mg/dL [3.0–3.8 mmol/L] and <54 mg/dL [<3.0 mmol/L]), and time above range (%TAR; the percentage of time with glucose levels 181–250 mg/dL [10.1–13.9 mmol/L] and >250 mg/dL [>13.9 mmol/L]). These data are presented in a color-coded figure that patients can easily understand. The report also displays an AGP graph that combines all of the glycemic values for the report period in a single 24-hour profile view, as well as individual glucose profile graphs for each day of the report period. In the following examples, we explain how to use these data to identify stage 1 and stage 2 dysglycemia.

For detection of early dysglycemia, it is important to consider all of the glucose data presented in the AGP report. For example, in Figure 1, the report of CGM data being reviewed at a 2- to 3-week follow-up visit, the area labeled A shows an average glucose value of 110 mg/dL, suggesting that the patient has normoglycemia based on findings from the study by Chan et al., in which an average glucose of 108 mg/dL was found to equate to an A1C of <5.7% (39). However, the Figure 1 daily glucose profiles labeled B show fairly frequent postprandial glucose excursions. This is an indication of loss of first-phase insulin response, which is often the first abnormality that occurs in early dysglycemia (43). The low %TAR shown in the section of the report labeled C suggests that the patient is still in the earliest stage of dysglycemia.

In the second example (Figure 2), we see a more elevated average glucose of 126 mg/dL in the section marked A, as well as glucose variability approaching 30% and a TAR of 3%. Correspondingly, we see significant postprandial hyperglycemia events occurring frequently in the daily glucose profiles marked B and a %TAR of 7% labeled C. These data suggest stage 2 dysglycemia and the need for more aggressive interventions, including both lifestyle change and possible pharmacologic therapy (40).

It is important to take some action at the follow-up visits. For individuals in stage 1, clinicians should emphasize lifestyle modifications such as transitioning to a healthier eating pattern, increasing physical activity, and losing weight. Clinicians may consider referring patients to a local Diabetes Prevention Program or to a registered dietitian or certified diabetes care and education specialist for support in these efforts. A CGM sensor can be started, necessitating another 2-week follow-up to discuss the data collected. Comorbid conditions (e.g., depression, hypertension, and hyperlipidemia) should be treated and/or referred to a specialist.

We recommend prescribing periodic CGM use (e.g., every 1–3 months) for individuals who express a desire to make such changes. More frequent CGM use may be needed if there are no improvements in glucose control.

For people in stage 2, it is important to explain why they face a greater risk of progression to type 2 diabetes and to give them an opportunity to comment first on what they want to focus on treating to reduce their cardiometabolic risk. Individuals may choose to focus on weight loss, and medications approved for this purpose by the U.S. Food and Drug Administration can be discussed. Referring to a printed page summarizing these medication options, including the expected approximate weight loss to be achieved and side effects, can assist this discussion (Table 3) (44–49).

A second option would be to follow ADA clinical guidelines for pharmacologic therapy focusing on comorbidities (50). For example, if a person has cardiovascular risk, a glucagon-like peptide receptor agonist may be considered; these agents can both help to address excess weight and dysglycemia and provide cardiovascular benefits (48). Later, the person may exhibit an early stage of kidney disease with an elevated UACR. If this is observed, treatment with a sodium–glucose cotransporter 2 inhibitor can be considered to address both dysglycemia and prevent or slow kidney disease progression.

When reviewing laboratory data, risk scoring, and CGM information, it is important to set a tone that empowers the patient; explaining the potential for progressing to stage 2 dysglycemia can set a patient’s expectations about the disease for years to come. Therefore, it is important to use the follow-up visit to provide information and tools to prevent the progression of the disease. It is also important to establish a collaborative relationship. Asking patients what they want to achieve from each subsequent appointment will help to build trust, which in turn can positively affect adherence to the care plan.

Be sure to explain what we know about the early pathophysiological changes associated with early dysglycemia that are already occurring in the body. An education aid such as a simple diagram of the human body can be used to explain and discuss these changes. For most people, it is best to start by explaining just two or three of these changes such as cardiovascular disease, kidney disease, and retinopathy. This strategy will provide a foundation to explain how lifestyle changes and medications might ameliorate these defects. Again, this discussion may be as simplified or as detailed as needed based on the patient’s interests and existing time constraints.

Periodic or continued use of CGM should be encouraged for individuals who are willing and financially able. However, it is important for those who continue to use CGM to routinely plan follow-up visits to discuss their glucose data. These discussions can occur in clinic visits or remotely via telehealth. Routine follow-up at 3- to 6-month intervals is recommended to reevaluate A1C, weight, CGM data (if continued), and other laboratory parameters.

Although numerous articles have reported the adverse clinical consequences of therapeutic inertia in managing type 2 diabetes (51), very little has been published regarding therapeutic inertia as it pertains to dysglycemia in its early stages. Given the increasing prevalence of overweight and obesity in the United States (1), clinicians increasingly will be challenged to identify individuals in these early stages and intervene early and effectively. Using multiple diagnostic tools and interventional strategies will provide the greatest opportunity for early identification of dysglycemia and prevention of clinical type 2 diabetes.

The authors thank Christopher G. Parkin of CGParkin Communications, Inc., for editorial support.

Funding for the development of this article was provided by Abbott Diabetes Care.

E.M. has received consulting fees from Abbott, Astra Zeneca, Boehringer Ingelheim, Eli Lilly, Merck, Novo Nordisk, and Sanofi US and has been a speaker for Abbott, Boehringer Ingelheim, Eli Lilly, and Novo Nordisk. K.M. has been a consultant for Novo Nordick and Semler. No other potential conflicts of interest relevant to this article were reported.

Both authors conducted literature searches, wrote and revised the manuscript, and approved the manuscript for submission. K.M. is the guarantor of this work and, as such, had full access to all the data presented and takes responsibility for the integrity of the data and the accuracy of the review.