Amid a subtle yet significant shift within diabetes care, continuous glucose monitoring (CGM) is emerging not simply as a tool for managing diabetes but as an aid to diagnose, predict and even prevent the condition. Diabetes remains one of the most underdiagnosed conditions globally. Just over four in ten people living with diabetes are undiagnosed, particularly in low- and middle-income countries. But even in high-resource settings, diagnostic blind spots exist.
CGM could be part of the solution. Unlike traditional methods such as the longstanding oral glucose tolerance test (OGTT)—a somewhat cumbersome and uncomfortable procedure, particularly for children—CGM provides a non-invasive, real-time window into changes in glucose levels over days and weeks. By continuously tracking blood glucose, a CGM device can reveal blood glucose patterns and fluctuations that a one-off test could easily miss. These patterns may not only signal an increased risk of developing diabetes, but—as newer research suggests—also foreshadow broader health complications.
With AI-driven analysis, CGMs can alert users to irregular glucose patterns so they can seek medical advice earlier and potentially reduce frequent medical appointments. When integrated with artificial intelligence and genetic risk profiling, CGM has the potential to transform personalised care and support the development of cellular therapies. In type 1 diabetes, these therapies replace damaged insulin-producing beta cells. CGM technology offers healthcare professionals a more accurate and dynamic method for evaluating transplant outcomes and measuring therapeutic effectiveness.
Given this potential, questions are emerging about whether current clinical thresholds for intervention are too low. As a result, researchers and clinicians are increasingly exploring CGM as a more accessible and potentially earlier diagnostic tool.
Unlike traditional methods such as the longstanding oral glucose tolerance test (OGTT), CGM provides a non-invasive, real-time window into changes in glucose levels over days and weeks
In pregnant women, the stakes are even higher. An estimated 1 in 5 live births in 2024 had some form of hyperglycaemia in pregnancy, while 1 in 6 are affected by gestational diabetes (GDM).
Gestational diabetes is often diagnosed late, between 24 and 28 weeks of pregnancy, and carries risks not just for the mother but for foetal brain development. Earlier screening and diagnosis would prompt CGM use as early as the 10th week of pregnancy, potentially bringing life-changing interventions for both mother and child.
At the recent International Diabetes Federation (IDF) World Diabetes Congress 2025, the scientific session “The Future of CGM” explored how CGM is transforming diabetes care—from its emerging role as a diagnostic tool to its impact on diabetes remission, cell therapies, and personalised interventions.
Professor Tadej Battelino of the University Medical Center Ljubljana, Slovenia, spoke of CGM use as a diagnostic tool across type 1, type 2 and gestational diabetes, stressing the potential of AI and machine learning in early detection and tracking disease progression. Professor Battelino advocated for CGM as a first-line option in prenatal care, offering a non-invasive and significantly more comfortable alternative to the traditional, sugar-heavy OGTT used to screen for gestational diabetes.
Among the implications of fluctuating blood glucose are far more troubling findings. A key study from the UK Biobank links glucose variability within the “normal” range to reduced brain volume and cognitive decline in people with type 1 diabetes. In fact, there is a link between changes in glucose patterns and cognitive decline, even in people whose levels are deemed ‘normal’ by current standards. Arguably, the damage begins even before the onset of diabetes.
However, the case for CGM is not merely technological—it is strategic. By pairing CGM data with genetic risk profiles and deploying AI to interpret the findings, clinicians could move from reactive treatment to predictive intervention. Yet this is not simply a matter of better gadgets. It is about redefining how we think about risk, diagnosis and responsibility. If glucose fluctuations harm the brain before diabetes is diagnosed, and if we can track and respond to those fluctuations at home, should we not act sooner?
By pairing CGM data with genetic risk profiles and deploying AI to interpret the findings, clinicians could move from reactive treatment to predictive intervention
As our understanding of type 1 diabetes grows, there is increasing emphasis on early detection and timely intervention. CGM is entering a new phase in managing the condition and identifying and assessing risk before symptoms appear. In the IDF 2025 scientific session, CGM use in pre-type 1 diabetes, Professor Chantal Mathieu, UZ Leuven, Belgium, explained how a combination of CGM, with genetic and immune profiling, the analysis of the immune system to understand its role in disease, could help redefine diagnostic criteria and support new strategies for early prevention.
For children newly diagnosed with type 1 diabetes, the traditional model has long prioritised insulin therapy and periodic monitoring via HbA1c levels. However, a wave of new research suggests a subtler picture that paints CGM as a diagnostic tool rather than only a blood glucose management device.
Using CGM data, researchers identified four clusters based on glucose management patterns. They introduced metrics such as “post-hypoglycaemic hyperglycaemia” (PHH)—sugar spikes following low episodes—as powerful indicators of risk or progression of the condition. The findings suggest that this temporary phase of improved glycaemic control—often called ‘remission’ or the ‘honeymoon phase’—is not merely a lucky lull but a metabolic state that is predictable with the right CGM thresholds.
In the domain of cutting-edge cellular therapies—such as islet cell transplantation and stem cell-derived insulin production—CGM is proving essential. Islet cell transplantation aims to restore natural insulin production in people with type 1 diabetes by using donor pancreas cells. Stem cell-derived insulin production offers a renewable and scalable alternative by generating insulin-producing cells in the lab, eliminating the need for donor organs.
As Professor Camillo Ricordi, University of Miami, USA, noted in his talk at the IDF Congress, CGM now informs not just candidate selection but also the real-time evaluation of graft success and early signs of rejection, allowing for timely clinical interventions.
Traditionally, transplant success was judged by whether a person with diabetes could stop insulin therapy. Now, more granular metrics such as “time-in-tight-range” (TITR) are closely related, more precise metrics than traditional “time-in-range” (TIR). TITR provides a more transparent lens into metabolic function.
CGM has also enabled breakthroughs in identifying candidates for immunosuppression-free autologous islet transplants, a procedure in which a person’s insulin-producing islet cells are harvested and re-implanted without needing immunosuppressive treatments. CGMs provide detailed, real-time insights into glucose patterns that reflect the body’s ability to regulate blood glucose. By doing this, people with the best chances of successful outcomes—and minimal need for lifelong medication—are selected.
In the domain of cutting-edge cellular therapies—such as islet cell transplantation and stem cell-derived insulin production—CGM is proving essential
The use of CGM in early-stage type 2 diabetes is equally transformative. Studies reveal that reliance on HbA1c or fasting glucose alone results in substantial underdiagnosis and misdiagnosis. In one population-level study, over 17% of people with diabetes were not diagnosed using traditional criteria—while others were incorrectly diagnosed with diabetes.
Long periods of high blood glucose—even slightly above normal levels—can be an early warning sign of intermediate hyperglycaemia (prediabetes) or type 2 diabetes. For example, spending more than 8–10% of the day with high blood glucose (above 140 mg/dL). This information, gathered from a CGM, could help healthcare professionals spot early changes in blood glucose before a complete diagnosis, giving people a chance to take action sooner.
Moreover, when combined with genetic risk factors and tests for specific autoimmune markers, CGMs can help build a much more personalised picture of a person’s risk of developing diabetes.
The economic implications are staggering. In many countries, diabetes-related costs account for a significant proportion of national health budgets. In 2024, global health expenditure related to diabetes reached an estimated USD 1.015 trillion, a 338% increase since 2007 and 12% of total global health spending. Estimates suggest that reducing “unhealthy lifespans” by one year globally could save $38 trillion. Such savings would come not only from diabetes prevention but also from a broader reduction in age-related condition, many of which share metabolic pathways with diabetes. The savings could be vast if CGM can help shift the model from reactive care to proactive prevention.
When combined with genetic risk factors and tests for specific autoimmune markers, CGMs can help build a much more personalised picture of a person’s risk of developing diabetes
As global consensus begins to form around CGM-based diagnostics, making the case for widespread adoption becomes harder to ignore. Yet this promising technology is still out of reach for many people with diabetes, particularly in low- and middle-income countries, were three in four of the estimated 589 million people currently living with diabetes reside.
Access to CGM devices remains severely limited due to high costs, inadequate healthcare infrastructure and shortages of trained professionals. Public awareness of early warning signs and prevention is low. In many cases, even basic diagnostic services are inaccessible. Without targeted investment and policy support for universal health coverage (UHC), the promise of CGMs may remain inequitable—delivering cutting-edge benefits to a few while leaving vulnerable populations behind.
Addressing these barriers will require a multi-pronged approach. Public health systems would benefit from bulk procurement and tiered pricing models with CGM manufacturers negotiated through governments and health agencies to reduce costs and improve affordability. Partnerships with non-governmental organisations and private companies could support the development of community-based training programmes to equip local healthcare workers with the skills needed to interpret CGM data and deliver culturally appropriate care. At the same time, investing in mobile health platforms could bridge infrastructure gaps by enabling remote monitoring and virtual consultations. Finally, public awareness campaigns, delivered through trusted community channels, can demystify CGM use and promote early screening, making prevention and proactive care a shared, accessible goal across populations.
The story is indeed changing. The widespread use of CGMs may result in earlier diagnoses, more personalised care and even delayed onset, reshaping clinical guidelines and the expectations of millions affected by or at risk of diabetes. Yet, meaningful progress in CGM-driven detection, prevention and management means ensuring access to this technology for the diabetes community worldwide.
