The research, published in The Lancet Regional Health – Europe, pooled data from the German/Austrian/Luxembourgian/Swiss Diabetes Patient Follow-up (DPV) Registry and from two Belgian real-world studies, the FUTURE and RESCUE trials, which together formed a joint Belgian cohort. In total, 4,178 adults with type 1 diabetes who started CGM—either alongside multiple daily insulin injections or insulin pumps—were followed for up to 24 months, alongside 2,104 propensity score-matched controls who did not use CGM. Matching on age, sex, diabetes duration, baseline weight and baseline HbA1c allowed the investigators to isolate, as far as possible in a retrospective design, the effect of CGM itself rather than the effects of simply living with type 1 diabetes through time.

On the surface, the average numbers seemed to confirm the old anxiety. In the DPV cohort, mean weight rose by 1.8 kg over 24 months after CGM initiation, accompanied by a small but statistically significant HbA1c reduction of 0.2 percent (about 1.7 mmol/mol). The Belgian cohort showed a more modest mean weight increase of 1.0 kg with a 0.07 percent HbA1c improvement. Even the matched controls, who never touched a CGM sensor, gained an average of 1.2 kg over the same period. These averages, however, told only part of the story. Real-world evidence had previously suggested an average body mass index increase of roughly 0.4 kg/m² after CGM initiation, and historical data from the landmark Diabetes Control and Complications Trial (DCCT) showed that intensive insulin therapy itself produced weight gains of 1.5 to 1.8 BMI units—a legacy that has coloured expectations of glucose-monitoring technologies ever since.

To move beyond averages, the team applied group-based multi-trajectory modelling, an application of finite mixture modelling that assumes the study population comprises a set of latent subgroups, each following a distinct polynomial trajectory over time, estimated by maximum likelihood. The number of groups and their polynomial orders were selected using the Bayesian information criterion, with a minimum group size of five percent of participants and average posterior probabilities of group membership exceeding 0.7. When applied simultaneously to the two outcome variables—body weight and HbA1c—the model revealed three clearly separable trajectories that replicated across all three cohorts. The first group, labelled “some weight gain with HbA1c decrease,” achieved the largest glycaemic improvements. The second and by far the largest group, “stable weight and HbA1c,” accounted for between 69.8 and 83.2 percent of participants depending on the cohort. The third and clinically most concerning group, “weight gain with stable HbA1c,” represented roughly one in ten CGM users.

The magnitude of change in these subgroups differed sharply. In the DPV-CGM cohort, individuals in the weight-gain group gained an average of 11.1 kg over 24 months while their HbA1c remained essentially unchanged. The Belgian counterpart gained 8.2 kg. Yet the most striking comparison emerged between CGM initiators and their matched controls: while only 9.9 percent of DPV-CGM participants and 10.4 percent of Belgian CGM users fell into the unfavourable weight-gain trajectory, the figure among matched controls without CGM was 20.5 percent—roughly double. Conversely, the proportion achieving the desirable combination of HbA1c improvement with limited weight gain was higher among CGM users. When the two CGM cohorts were pooled and analysed through a multivariate random-effects meta-analysis with hierarchical cohort-level random intercepts, the pooled estimates placed 77.8 percent of CGM initiators in the stable trajectory and only 9.6 percent in the substantial-weight-gain trajectory.

Who ends up on the adverse path? Multinomial logistic regression, using the stable group as reference, identified early weight dynamics as the dominant predictor. Every additional kilogram gained above the cohort average during the first six months after CGM initiation nearly doubled the odds of belonging to the weight-gain-with-stable-HbA1c subgroup, with odds ratios of 1.85 in the DPV-CGM cohort, 1.91 in the Belgian cohort and 1.98 in the combined DPV analysis. Baseline characteristics mattered as well: those in the adverse subgroup were heavier at the outset, with the highest baseline weight and the largest proportion of participants in the obesity category (BMI ≥30 kg/m²), and they tended to be younger with shorter diabetes duration. Meanwhile, individuals in the metabolically favourable subgroup—who lost 2.3 to 2.4 percentage points of HbA1c in the DPV cohort—started with the highest baseline HbA1c (9.6 percent in DPV-CGM) and the lowest baseline weight, and each percentage point of HbA1c above the cohort average reduced the odds of belonging to that group by roughly 95 percent, underscoring how much headroom for improvement distinguished this population.

Perhaps the most provocative finding came from the direct comparison of matched populations within the DPV registry. In a multinomial model fitting both CGM users and controls, individuals who initiated CGM were almost three times more likely (odds ratio 2.80, 95 percent confidence interval 1.56–5.04) to belong to the group achieving substantial HbA1c reductions without excessive weight gain than their non-CGM counterparts. The fitted models showed strong explanatory power, with Nagelkerke pseudo R² values between 0.58 and 0.66. Complementary cross-lagged panel analysis, which examines bidirectional influences between two variables across repeated time points, found that weight and HbA1c largely predicted their own future values, with little cross-variable coupling—except in the adverse DPV-CGM subgroup, where higher HbA1c at 12 months predicted greater weight at 24 months, at a rate of nearly one additional kilogram per percentage point of HbA1c above the mean. This hints at a dynamic interplay between deteriorating glucose control and accelerating weight gain in vulnerable patients.

The clinical interpretation offered by the researchers is twofold. First, the notion that CGM inherently causes weight gain does not survive scrutiny when individual trajectories rather than population averages are examined. Prior real-world reports of an average BMI increase of about 0.4 kg/m² after CGM initiation—equivalent to roughly 1.1 kg in a person of 1.7 metres—are, in this analysis, driven largely by a small minority, while the substantial majority maintain stable weight and glycaemia. Second, and more urgently actionable, the first six months after CGM initiation constitute a critical early-warning window. A patient who gains approximately 1.5 kg in that period nearly doubles their risk of joining the subgroup that ultimately gains 8 to 11 kg over two years. The authors argue that clinicians should treat early weight trajectory as a red flag and consider timely interventions for this subset—potentially including adjunctive weight-management pharmacotherapy or, in selected cases, referral for metabolic surgery—to prevent the emergence of what the field calls “double diabetes,” a phenotype combining the insulin deficiency of type 1 diabetes with the insulin resistance and central adiposity of type 2 diabetes, and with it an elevated risk of cardiovascular and microvascular complications.

The study’s design carries both strengths and caveats worth weighing. Its reliance on robust longitudinal real-world data from three European sources lends ecological validity that randomised trials, with their tightly protocolised conditions and shorter horizons, often lack. The consistency of the three-trajectory structure across independently collected cohorts, confirmed formally by meta-analytic pooling, strengthens the case that these patterns reflect genuine biology and behaviour rather than statistical artefact. Yet the retrospective design leaves open the possibility of unmeasured confounding, including social determinants of health, physical activity and dietary habits that could shape both weight and glucose trajectories. The predominantly European cohorts, with limited ethnic diversity and similar healthcare structures, may constrain generalisability. Moreover, most participants used multiple daily injections rather than insulin pumps, and fewer than two percent were on hybrid closed-loop systems, which have demonstrated glycaemic benefits beyond what CGM paired with injections can achieve—making evaluation of weight outcomes under automated insulin delivery an obvious priority for future research.

For the millions of adults with type 1 diabetes weighing whether to adopt CGM, and for the clinicians advising them, the message is clearer than it has ever been. Continuous glucose monitoring, by replacing episodic fingerstick snapshots with a continuous stream of interstitial glucose data, enables finer insulin titration, better detection of overnight and postprandial excursions, and generally more confident self-management—all without, for the great majority of users, a penalty in body weight. Indeed, the data suggest the technology may be part of the solution to the growing problem of overweight and obesity in type 1 diabetes, since people without CGM were twice as likely to drift onto a substantial weight-gain path. What remains is a shift in clinical practice: away from population averages and toward personalised monitoring of individual trajectories, with early weight trends after CGM initiation serving as a simple, inexpensive signal of who needs intensified metabolic support before excess weight becomes entrenched.

Continuous Glucose Monitors have revolutionized diabetes management by providing real-time blood glucose readings, facilitating proactive treatment adjustments. However, the authors argue that despite their efficacy, there exists a significant inconsistency in how these devices are prescribed across various demographics and healthcare systems. Their study emphasizes that access to CGMs is not uniformly distributed, raising concerns about fairness and opportunity in patient health management.

The researchers leveraged extensive databases and analyzed prescription trends among a diverse cross-section of primary care physicians. The findings reveal that socio-economic factors significantly influence CGM prescribing patterns. They discovered that patients from lower-income backgrounds, as well as those residing in rural areas, frequently face barriers to accessing necessary medical technologies. Such systemic inequities put vulnerable patient populations at a higher risk for diabetes-related complications.

Moreover, this inequity extends beyond economic factors. The study uncovered variations linked to racial and ethnic backgrounds, suggesting that certain groups are disproportionately affected by the lack of CGM prescriptions. The authors pushed for a more nuanced understanding of healthcare disparities, emphasizing that measures must be taken to address both the socio-economic and cultural dimensions of healthcare access.

The implications of such findings are staggering. Without intervention, these inequities will only widen, potentially resulting in increased morbidity and mortality rates among populations already burdened by chronic diseases. In response to these disparities, the authors advocate for policy reforms that prioritize equitable access to diabetes technologies, including CGMs. Implementing such changes may ensure that all patients, irrespective of their background, receive optimal care.

Furthermore, the authors suggest that healthcare providers must undergo training to recognize and mitigate biases in prescribing practices. By fostering awareness of these disparities, the healthcare community can work collaboratively towards more equitable practices that prioritize patient health above all. Advocating for inclusive policies will not only enhance patient well-being but also improve overall public health outcomes.

In light of the findings, stakeholders in the healthcare sector are called to action. Policymakers must recognize the importance of balancing technological advancements with equitable distribution, ensuring that all patients can benefit from innovations such as CGMs. Additionally, primary care physicians are urged to engage in ongoing education about the importance of equitable prescribing practices.

The role of technology in healthcare cannot be overstated, yet access to such resources must remain a priority. As this study elaborates, the goal of modern medicine is not only to treat but also to empower patients through accessible and equitable healthcare solutions. Addressing the inequities in CGM prescribing behaviors paves the way for a brighter, healthier future.

In conclusion, the work of Milosavljevic and colleagues challenges us to reconsider our healthcare systems’ structure and practices. By exposing the disparities in CGM prescriptions, they have laid the groundwork for critical discussions on equity in healthcare. It is paramount that both healthcare professionals and policymakers heed these findings—acting decisively to create a future where every diabetic patient can monitor their glucose levels effectively and obtain the care necessary to live healthier lives.

The ramifications of these findings go beyond mere academic discourse; they resonate deeply within the hearts of families and communities affected by diabetes worldwide. With over 463 million adults currently living with the disease, the need for equitable healthcare practices has never been more urgent. The urgency to act by implementing supportive policies is clear, ensuring CGM and other essential medical resources reach those who need them most.

Both hope and change rest heavily in the hands of the healthcare system, which must shift towards practices that prioritize equity, inclusivity, and comprehensive patient care. This is no longer just a suggestion but a necessity in our increasingly complex and diverse world of medicine. Each step toward equitable healthcare practices brings us closer to a world where diabetes management is a right—a foundation upon which healthier societies can be built.

Subject of Research: Inequity in Continuous Glucose Monitor (CGM) prescribing behaviors in primary care
Article Title: Inequity in Continuous Glucose Monitor (CGM) Prescribing Behaviors in Primary Care
Article References: Milosavljevic, J., Schechter, C., Fazzari, M. et al. Inequity in Continuous Glucose Monitor (CGM) Prescribing Behaviors in Primary Care. J GEN INTERN MED (2026). https://doi.org/10.1007/s11606-025-09923-7
Image Credits: AI Generated
DOI: https://doi.org/10.1007/s11606-025-09923-7
Keywords: Inequity, Continuous Glucose Monitoring, Diabetes, Healthcare Access, Primary Care, Prescription Behaviors