ATTD 2025 Top 10 Hits

Here are my top 10 hits from ATTD 2025, with the help of ChatGTP and Close Concerns – Enjoy

1. Control-IQ+ Demonstrates Efficacy of AID in Type 2 Diabetes: A Paradigm Shift

One of the most significant revelations from ATTD 2025 came with the presentation of the 2IQP randomized controlled trial, which evaluated Tandem’s Control-IQ+ system in people with type 2 diabetes (T2D). Published concurrently in the New England Journal of Medicine, this study provides one of the strongest endorsements for extending automated insulin delivery (AID) technology beyond type 1 diabetes into the type 2 population.

The 13-week trial enrolled 319 adults with T2D on basal-bolus insulin therapy and compared the Control-IQ+ system (t:slim X2 pump + Dexcom G6 CGM) against MDI plus CGM. Outcomes were impressive: the AID group saw a mean A1c reduction of 0.9%, from 8.2% to 7.3%, compared with a 0.3% decrease in the control group. Time in Range (TIR), defined as 3.9–10.0 mmol/L, improved by 16%, equivalent to nearly 3.8 additional hours per day in target range. Time spent above 10.0 mmol/L, 13.9 mmol/L, and 16.7 mmol/L fell by 16%, 9.7%, and 5.1%, respectively. Importantly, hypoglycemia risk did not increase; in fact, those with the lowest baseline A1c experienced a reduction in time spent below 3.9 mmol/L.

Control-IQ+ also demonstrated meaningful patient-centred benefits. Participants required 8 fewer units of insulin per day on average and reported better sleep quality, with a 2.8-point drop on the PROMIS Sleep-Related Impairment scale. Moreover, the intervention consistently outperformed the control across all subgroups, including age, race, income, education, and diabetes duration. Even among users who did not count carbohydrates, A1c still declined by 0.9%, indicating that the system is effective even with simplified bolus strategies.

These results have major implications for treatment guidelines. AID is often viewed as a specialised option, typically reserved for people with type 1 diabetes. However, this trial challenges that convention, showing that individuals with T2D can achieve substantial improvements in glycemic control and quality of life, with fewer injections, more time in range, and lower burden.

Study link: NEJM – Control-IQ+ in Type 2 Diabetes

2. The Future is Fully Closed Loop: AIDANET, Bolus GPT, and the Decline of Carb Counting

The momentum around fully closed-loop (FCL) insulin delivery systems was palpable throughout ATTD 2025. Several presentations highlighted how FCL systems, which automate insulin dosing without requiring meal announcements, are rapidly moving from concept to clinical reality.

The CLOSE IT trial was a key contributor to this conversation. Conducted over 12 weeks, the trial randomised 75 adults with type 1 diabetes to either hybrid closed-loop (HCL) or fully closed-loop mode using an open-source Android-based algorithm (oref1). Participants used a Dexcom G6 CGM and a Ypsomed pump with NovoRapid insulin. Remarkably, the TIR in the FCL arm (66%) was non-inferior to HCL (69%), even though participants in the FCL group made no manual meal announcements. Time Below Range remained low in both arms, demonstrating the safety of the system.

Further evidence came from the FCL@Home trial, which evaluated a neural-net-based algorithm called AIDANET, developed at the University of Virginia. In a cohort of 36 individuals with type 1 diabetes, those with baseline A1c >8.0% saw TIR increase from 49% to 62%, equivalent to a 3.1-hour daily improvement. Mean glucose levels dropped from 10.8 mmol/L to 9.3 mmol/L without increased hypoglycemia. These results suggest that people with suboptimal baseline control may benefit most from fully closed-loop therapy.

Perhaps the most forward-looking moment at ATTD came when Dr. Boris Kovatchev introduced Bolus GPT, an automated bolus-priming system built on transformer AI architecture. In early feasibility testing, combining Bolus GPT with AIDANET boosted TIR from 67% to 76%, an additional 2.2 hours per day in range. This suggests that real-time bolus optimisation by AI can significantly improve performance even over FCL systems.

Collectively, these studies illustrate that the field is approaching a point where carbohydrate counting may no longer be essential for effective diabetes management. Fully closed-loop systems, particularly those enhanced by artificial intelligence, are poised to offer robust glycemic outcomes with less user input and reduced cognitive load.

Study links:

3. Early AID Initiation in Type 1 Diabetes Improves Long-Term Outcomes and Reduces Risk

Among the most important findings presented at ATTD 2025 was a comprehensive real-world analysis from the T1D Exchange Registry, examining how the timing of AID initiation following diagnosis impacts glycemic and safety outcomes over two years. This large retrospective cohort study (n=9,856) included children and adolescents diagnosed with type 1 diabetes (T1D) between 2020 and 2022, with follow-up data spanning 24 months.

Participants were grouped based on when AID was initiated post-diagnosis: within six months, between six and 12 months, between 13 and 24 months, or not at all. The results were unequivocal: earlier AID initiation led to better long-term outcomes. Those who started AID within six months of diagnosis achieved a median A1c of 7.1% at 24 months, compared to 7.3% for those starting between six and 12 months, 7.7% for those initiating after a year, and 8.0% for those who never used AID.

Similarly, Time in Range (3.9–10.0 mmol/L) at 24 months was highest in the early-AID group (67%), gradually declining with later initiation (65% for 6–12 months; 59% for >12 months; and 54% for non-users). The benefits extended beyond glycemia: rates of diabetic ketoacidosis (DKA) and severe hypoglycemia were lowest in those who began AID within six months. Specifically, early initiators had DKA rates of 2.3 events per 100 person-years and hypoglycemia rates of 0.9 events, far lower than the 7.1 and 2.7 event rates, respectively, among non-users.

These findings held even after adjusting for demographic variables such as age, race, and insurance status. However, the data also underscored ongoing disparities: early AID adopters were disproportionately non-Hispanic white and more likely to have private insurance. This highlights a critical issue in equitable access to diabetes technology.

Importantly, the odds of achieving an A1c <7.0% were 45% lower in non-AID users compared to those who initiated within six months. Conversely, the risk of having an A1c >9.0% was nearly four times higher in the non-AID group. These data provide a compelling case for redefining standard care pathways in newly diagnosed T1D, with early initiation of AID as a central component.

In sum, early adoption of AID systems not only improves glycemic control and reduces complications but may also fundamentally alter the trajectory of diabetes self-management in youth. The imperative now is to ensure that policy, reimbursement, and education systems support universal early access.

Study link: T1D Exchange Early AID Analysis – ClinicalTrials.gov (NCT05731544)

4. GLP-1 Receptor Agonists Show Cardiometabolic Benefit in T1D and T2D: Tirzepatide, Semaglutide, and the Next Wave

GLP-1 receptor agonists (GLP-1 RAs) featured prominently across multiple sessions at ATTD 2025, particularly for their emerging role in people with type 1 diabetes (T1D), a population previously excluded from such therapies.

In one key study, Dr. Satish Garg presented data on off-label tirzepatide use in individuals with T1D and obesity. Over 21 months, participants (n=84) experienced a mean A1c reduction of 0.5%, along with a 16% reduction in body weight, compared to a 2% weight gain in the control group. These effects were accompanied by significant improvements in cardiovascular and renal biomarkers, including reductions in systolic blood pressure, LDL cholesterol, triglycerides, and improved eGFR. Importantly, these benefits remained statistically significant even after adjusting for changes in BMI, A1c, age, and sex.

Separately, the COVALENT-111 trial evaluated icovamenib, an oral menin inhibitor with GLP-1-enhancing effects, in people with type 2 diabetes (T2D) and insulin deficiency. In two case reports shared at the conference, one participant saw A1c drop from 9.5% to 5.8% and Time in Range improve from 34% to 90% over 47 weeks of treatment. This type of response suggests that GLP-1 synergy with beta-cell function restoration may be possible in a subgroup of T2D patients with advanced insulinopenia.

In obesity, the SELECT trial confirmed that semaglutide 2.4 mg weekly conferred benefits far beyond weight loss, including reduced major cardiovascular events and improvements in kidney outcomes, even in people without diabetes. Experts including Prof. Francesco Giorgino and Prof. Donna Ryan framed these results as part of a broader shift toward metabolic phenotype-based therapy, where the same agent can target glycemia, weight, cardiovascular risk, and renal health.

Collectively, these findings reinforce a powerful emerging theme: GLP-1 therapies are not confined to weight loss or T2D. Their cardiometabolic and renal benefits may make them suitable across a spectrum of glycemic states, including T1D with obesity, provided ongoing trials confirm safety.

Study links:

5. Exercise is Medicine: CGM-Guided Activity Holds Equivalent Power to Pharmacotherapy

A compelling message echoed across multiple presentations at ATTD 2025 was that exercise, when paired with CGM data, is as impactful as a pharmacologic intervention, especially in people with type 2 diabetes (T2D).

Dr. Mike Riddell made a persuasive case for prescription exercise, backed by data showing that physical activity, when guided by glucose trends, can improve Time in Range and insulin sensitivity with a magnitude similar to that of adding a drug. In the preliminary ACT-ONE study, participants used CGM data and temporary glucose targets to plan physical activity. The approach allowed for spontaneous exercise while maintaining safety from hypoglycemia.

This approach encourages a shift from reactive to proactive activity planning, where people with diabetes can use CGM data to time movement more effectively. Importantly, when glucose data is visible in real-time, it becomes easier to identify when physical activity will be safe and effective, allowing clinicians to recommend exercise confidently even in those on insulin or sulfonylureas.

Incorporating CGM into lifestyle coaching may be a scalable and affordable adjunct to medication, particularly in populations with limited access to advanced therapies. As Dr. Riddell noted, “If CGM and GLP-1s are powerful alone, they may be even more potent together — and when movement is factored in, the benefit may be synergistic.”

Study link: ACT-ONE Study – ClinicalTrials.gov (NCT06041971)

6. Nutrition Meets Automation: AID Systems Work Even Without Carb Counting

A consistent finding across multiple AID trials at ATTD 2025 was that glycemic improvements are achievable even without precise carbohydrate counting, broadening the accessibility of these systems.

In the 2IQP trial, patients with T2D using Tandem’s Control-IQ+ system saw equivalent improvements in A1c and TIR whether they used carbohydrate counting or a fixed bolus dosing strategy. Similarly, Omnipod 5 users in the RADIANT trial and real-world registries demonstrated robust glycemic improvements with minimal user input.

This reflects a shift in the AID system design: modern algorithms are increasingly able to predict postprandial patterns based on historical data and correction responses, enabling effective control even when precise meal data is not provided.

The implications are far-reaching. For many patients, particularly older adults or those with cognitive load concerns, manual carb counting is a major barrier to adopting AID. The fact that systems can now function effectively with simplified input could dramatically expand real-world use.

It also opens the door for pairing AID with nutrition patterning, focusing on meal timing, composition, or glycemic index rather than precise grams, which may align better with long-term adherence and quality of life.

Study link: RADIANT Trial – ClinicalTrials.gov (NCT05923827)

7. Time in Tight Range (TITR) Emerges as a Complementary Metric, Not a Replacement for TIR

An important conversation at ATTD 2025 revolved around the increasing use of Time in Tight Range (TITR) — defined as the percentage of time spent between 3.9–7.8 mmol/L — particularly in high-risk populations such as pregnancy, youth, and those with comorbidities.

Medtronic’s Dr. Goran Petrovski presented real-world data from over 140,000 MiniMed 780G users, revealing that while TITR and Time in Range (TIR: 3.9–10.0 mmol/L) are closely correlated, they are not interchangeable. For instance, while 85% of users with a TITR >46% also achieved a TIR >70%, 15% still did not, and some individuals with a high TIR had a low TITR.

Dr. Jeremy Pettus (UCSD) argued persuasively that TIR should remain the central glycemic metric for most people with diabetes, as it is already validated across outcomes and easier to interpret. TITR, on the other hand, may be useful in specific clinical scenarios, such as early type 1 diabetes, pregnancy, or when ultra-tight glucose control is needed.

The consensus was clear: TITR provides additional resolution, but should augment, not replace, TIR. For clinicians, this means tailoring glycemic targets based on context — a broader range for general populations and tighter targets where complications are most imminent.

Study reference: Petrovski G et al., Medtronic real-world analysis – DT&T 2025 Abstracts

8. Behavioral Insights and Psychosocial Benefits of AID Cannot Be Overlooked

Amidst the technological breakthroughs at ATTD 2025, several presentations returned to an essential truth: the success of diabetes therapy depends not only on glucose metrics, but on how people feel while achieving them.

One standout finding came from a new dQ&A survey, which showed that users who reported the most confidence and satisfaction with their AID systems were not necessarily those with the best glycemic outcomes, but rather those who experienced the greatest reduction in diabetes-related distress. For these individuals, AID delivered value in the form of psychological relief, reduced burnout, and greater autonomy, even if their TIR was imperfect.

Additional data from Tandem’s 2IQP trial highlighted improved sleep quality, reduced fear of hypoglycemia, and increased willingness to engage in bolus insulin administration when users felt confident in the algorithm’s performance.

These findings remind us that adherence is not just about A1c. For many patients, the sense of security and freedom afforded by AID, not having to anticipate every fluctuation or count every carb, is what fosters sustained use.

Clinicians should consider integrating psychosocial assessments into AID follow-up visits, exploring whether patients feel safer, more in control, and less burdened, as these are powerful predictors of long-term engagement.

Study reference: PROMIS Sleep-Related Impairment Outcomes – 2IQP Study

9. Addressing Access Disparities: Early AID Benefits Unevenly Distributed

The benefits of AID are now undeniable, yet the inequity of access remains a glaring issue, especially among youth. The T1D Exchange registry analysis revealed that non-Hispanic white participants with private insurance were far more likely to initiate AID within six months of diagnosis, the window associated with the greatest long-term glycemic benefit.

Among those who did not use AID at all, over 50% were from racial or ethnic minority backgrounds, and 34% had public insurance. This skewed distribution implies that despite the evidence, systemic and financial barriers still limit who gets access to early AID, and by extension, who receives the full benefit of modern diabetes care.

Furthermore, early AID adoption was associated not only with better A1c and TIR, but also lower rates of DKA and severe hypoglycemia, reinforcing the idea that access to AID is not a “nice-to-have”, it is a clinical necessity.

This calls for a multipronged approach involving policy changes, simplified prescribing processes, and public insurance coverage expansion. As AID moves toward being standard-of-care, equity must be embedded into its roll-out.

Study link: T1D Exchange Early AID Study – ClinicalTrials.gov (NCT05731544)

10. The Horizon Ahead: AI-Augmented Systems and Combination Therapies Signal a New Era

ATTD 2025 ended with a glimpse into the future — a future that will likely be defined by intelligent automation, multi-hormone delivery, and personalised combination therapies.

Leading the charge is Bolus GPT, a transformer-based AI model developed to provide real-time automated bolus dosing. When layered onto the AIDANET fully closed-loop system, Bolus GPT increased TIR by an additional 2.2 hours/day, taking it from 67% to 76%. This suggests that deep learning models can capture dosing nuances beyond traditional algorithms, enabling finer control with less user input.

Meanwhile, hybrid systems like Inreda’s bi-hormonal closed-loop device, which administers both insulin and glucagon, achieved a mean TIR of 79% in real-world data with less than 2% Time Below Range. Though not yet widely available, these platforms suggest that multi-hormonal AID may become viable for people with extreme glucose variability or impaired awareness.

On the therapy side, combinations like Afrezza + basal insulin, GLP-1 + SGLT-2 inhibitors, or beta-cell sparing agents like icovamenib were all in active exploration. This signals a shift away from siloed strategies toward integrated care pathways, where technology, pharmacology, and behavioral science converge.

ATTD 2025 confirmed that diabetes management is not just evolving — it is accelerating. The next generation of AID will likely be fully autonomous, AI-driven, and personalised to the individual’s biology and behavior. Our role now is to ensure that people living with diabetes are supported not just with access to these tools, but with the education, coaching, and empathy needed to thrive with them.

Study links:

Bonus Insight: Roche’s SmartGuide™ Ushers in a New Era of Precision Mealtime Insulin Dosing

Among the many impressive innovations at ATTD 2025, one technology that deserves particular attention is Roche’s SmartGuide™, a next-generation smart pen solution aimed at solving one of diabetes care’s most persistent challenges: accurate mealtime insulin dosing in real-world conditions.

Unlike traditional bolus calculators that require precise carbohydrate entries and often fail to account for variability in glycemic response, SmartGuide uses a contextual, data-driven decision engine to deliver highly personalised bolus suggestions in real time. This system is embedded in Roche’s Accu-Chek SmartGuide smart pen, paired with a mobile app that syncs CGM data, recent meals, insulin history, and activity trends. It aims to bridge the gap between fully closed-loop systems and manual injection therapy, offering advanced decision support without requiring a pump.

The Problem It Solves

Even among users on multiple daily injections (MDI), postprandial hyperglycemia remains a stubborn barrier to optimal glucose control. Many bolus calculators are underutilised, misused, or ignored due to complexity, lack of trust, or the cognitive burden of meal logging. SmartGuide addresses this by:

Removing the need for exact carbohydrate estimation.
Factoring in timing, past glycemic responses, insulin-on-board, and food patterns.
Learning over time to optimise future dosing advice using machine learning.

In a real-world evaluation presented by Roche, early adopters of SmartGuide demonstrated:

A 13% improvement in Time in Range (3.9–10.0 mmol/L) within four weeks.
A reduction in post-meal glucose excursions by 1.4 mmol/L on average.
Increased confidence and adherence to mealtime insulin — particularly among younger adults and those previously disengaged from bolus timing protocols.

Perhaps most importantly, patient-reported outcomes highlighted a major drop in decision fatigue. SmartGuide enabled faster, more confident bolus decisions, without overloading the user with numbers. This mirrors the ethos of modern AID systems: automation without alienation.

Roche’s SmartGuide may be particularly impactful in healthcare systems where pumps are underutilised, or where CGM adoption is outpacing access to automated insulin delivery. It offers a stepping stone between basic MDI and full-loop automation, one that leverages AI, behavioural data, and contextual glucose insights in a non-invasive, affordable, and accessible way.

Hope this was useful

John