TL;DR
New international guidance is now available on how to exercise safely with type 1 diabetes when using Automated Insulin Delivery (AID) systems.
- The EASD–ISPAD consensus statement summarises the best current evidence on AID, exercise, and glucose management.
- Key levers for safer exercise: adjust glucose targets, manage carbohydrate around activity, minimise insulin on board, and use each system’s specific activity modes.
- Downloadable graphics are available in both mmol/L and mg/dL, plus one-page TL;DR guides for major commercial systems.
- This page is information, not individual medical advice. Discuss any changes with your own diabetes team.
The simple version
It’s finally landed.
The Automated Insulin Delivery (AID) and Exercise Consensus Statement pulls together what we currently know about exercising safely with type 1 diabetes when using AID systems.
You can download all the key graphics here:
mmol/L graphics (PowerPoint)
and
mg/dL graphics (PDF).
Short, system-specific TL;DR guides are available here:
This page walks through the core ideas, some system-specific nuance, and practical actions you can discuss with your team and experiment with in real life.
The medium version
As a member of the writing group for the European Association for the Study of Diabetes (EASD) and the International Society for Pediatric and Adolescent Diabetes (ISPAD) position statement, I am honoured to have contributed to this guidance.
The consensus document consolidates the latest evidence and clinical practice into a single resource aimed at people with type 1 diabetes, families, and healthcare professionals. The goal is simple: safer, more enjoyable movement with AID systems, using practical tools that can be implemented in day-to-day life.
Here are the people to thank, with special appreciation to Othmar and Dessi who drove the consensus process (and were probably driven slightly mad in the process).
A little background
Exercise is one of the most powerful tools for managing type 1 diabetes, improving cardiovascular health, insulin sensitivity, mental health, and quality of life. Yet, when you add an AID system into the mix, things can get complicated.
AID systems adjust insulin frequently in the background. This is brilliant for overnight and day-to-day living, but during exercise it means you start each session with different “insulin conditions”, even when your starting glucose looks the same. That makes trial-and-error hard and repeatability low.
This consensus statement lays out a framework for thinking about exercise with AID: how to shape insulin delivery before, during, and after activity; how to use system-specific tools; and how to make the whole process safer and more predictable.
If you are completely new to exercise and type 1 diabetes, here is a starter guide that covers the basics before you add AID into the picture.
Core principles for safe exercise with AID systems
The consensus outlines key strategies that address the main moving parts of physical activity and glucose management with AID systems.
-
Planned physical activity
When you expect glucose to fall, increasing the glucose target 1–2 hours before exercise is usually helpful. For activities that tend to raise glucose (e.g. short, very high-intensity work), keeping regular or even slightly lower targets may be more appropriate.If exercise starts within 2 hours of a meal, reduce the carbohydrate amount entered into the AID system by around 25–33%. This reduces bolus insulin, lowering hypoglycaemia risk, even if glucose temporarily runs a little higher. The key point: the reduction in carbs means a net reduction in insulin; the algorithm cannot “recreate” the insulin you never gave.
-
Unplanned physical activity
For sudden activity (e.g. a spontaneous game, unexpected walk), adjust in real time by setting a higher glucose target and using small amounts of rapid-acting carbohydrate if glucose is below about 7.0 mmol/L (126 mg/dL) and trending down. -
Carbohydrate management
Balanced carbohydrate intake around exercise aims to prevent hypoglycaemia without causing prolonged hyperglycaemia afterwards. A useful starting point is to begin supplementing with carbohydrate below 7.0 mmol/L (126 mg/dL), using roughly 3–20 g every 20–30 minutes. The exact amount depends on trend arrows and what you expect your glucose to do during the session.The consensus includes one of my favourite graphics of this kind, which helps you think about likely glucose behaviour during exercise.
-
Timing and insulin on board (IOB)
Exercising with less active insulin on board reduces hypoglycaemia risk. A pragmatic “three-hour rule” is to avoid starting planned exercise within three hours of your last significant bolus if possible, or to specifically reduce that bolus if you will be active.Planning activity before meals or during fasting windows is often safer and more predictable.
-
Device adaptations
Most AID systems include temporary modes or targets designed for activity. Used well, these features can reduce hypoglycaemia risk and smooth glucose swings during and after exercise.
These strategies emphasise understanding your own glucose responses, then layering AID features on top of that understanding rather than relying on the algorithm alone.
System-specific considerations
The position statement provides tailored guidance for different AID systems, supported by excellent system-specific infographics created by Othmar’s wife.
Beta Bionics iLet
Key point: iLet has limited ability to adjust glucose targets before exercise, so carbohydrate planning and real-time carb supplementation become particularly important.
CamDiab mylife CamAPS FX
CamAPS FX offers an “Ease-off” mode that raises glucose targets and makes the algorithm less aggressive, reducing hypoglycaemia risk during activity. This can be combined with increasing the personal glucose target where needed.
Medtronic MiniMed 780G
MiniMed 780G uses a “Temp Target” setting that raises the target glucose and stops automated correction boluses during exercise. This can be particularly useful for preventing repeated hypoglycaemia with prolonged or aerobic activity.
Omnipod 5
Omnipod 5 includes an “Activity” feature that sets a higher glucose target 1–2 hours before exercise and makes the SmartAdjust algorithm less aggressive. Used in advance, this can reduce the amount of insulin on board at the start of activity.
Tandem t:slim X2 Control-IQ
Tandem’s “Exercise Mode” adjusts the target range upward and moderates insulin delivery, supported by predictive algorithms to reduce hypoglycaemia risk while still managing post-exercise rises.
DBLG1 system
The DBLG1 system includes a “Physical Activity” mode for exercise and a “ZEN” mode that increases the target glucose level to offer additional protection against lows.
- DBLG1 system: Use “Physical Activity” mode during exercise; consider ZEN mode when you want a higher protective target (for example, for long or unpredictable sessions).
Each system has its own features and limitations. The art is in tailoring strategies to your preferences, your physiology, and your device’s behaviour.
Practical guidance for different types of exercise
Different forms of activity produce different glucose responses. Broadly:
- Aerobic activities (e.g. steady-state running, cycling, swimming) typically lower glucose during and shortly after exercise.
- High-intensity or sprint intervals may initially raise glucose because of counter-regulatory hormones, with a later drop once those hormones settle.
- Resistance training can produce variable patterns, often with smaller immediate drops but meaningful longer-term improvements in insulin sensitivity.
Some practical patterns from the consensus:
- Aerobic activities: Start with a higher glucose target if possible and use small, frequent carbohydrate doses to stay in range. Using exercise/activity modes in advance is particularly helpful here.
- High-intensity exercise: Consider a modest reduction in insulin doses beforehand if you are prone to delayed lows, and monitor closely during the hours after the session to catch late drops.
- Prolonged activities: For longer sessions (e.g. hikes, long bike rides, team matches), additional planning is needed. Carry fast-acting carbohydrate, consider glucagon where appropriate, and be prepared for both hypo- and hyperglycaemia, especially when adrenaline and fatigue interact with the algorithm.
Overcoming challenges and implementing recommendations
Even with AID, exercise is rarely perfectly smooth. A few persistent challenges tend to show up:
- CGM lag: Interstitial glucose lags behind blood glucose, particularly during rapid changes. Trend arrows and the direction of recent movement often matter more than any single number.
- Algorithm behaviour: AID systems respond to what they “see”. A sharp rise before exercise may prompt extra insulin, which then increases hypoglycaemia risk once you start moving.
- Individual variability: The same session can produce different glucose patterns on different days depending on sleep, stress, recent activity, illness, and hormonal status.
To navigate this, the consensus encourages people to:
-
Work closely with their healthcare team to understand their own glucose trends and risk patterns.
-
Use CGM trend arrows and system data to guide real-time decisions, rather than reacting only to single values.
-
Regularly review exercise sessions, adjust strategies, and treat each block of activity as another experiment that refines your personal playbook.
When AID makes exercise feel unpredictable
Struggling with erratic glucose levels around exercise?
This is a common experience with AID systems. Because the algorithm adjusts insulin every few minutes, you effectively never start a session with exactly the same insulin profile twice. That makes pattern recognition and planning much harder.
One strategy described in clinical practice is to temporarily step out of full automation to increase reproducibility:
- Set a manual basal rate appropriate for your planned exercise.
- Switch to open-loop (or a less automated mode) around two hours before starting the session.
- Keep this stable basal running into and during the activity, then use small carbohydrate adjustments as needed.
This approach stabilises insulin conditions so that your responses from session to session become more reproducible. Over time, it becomes easier to understand how much carbohydrate you personally need for a given type and duration of activity.
After exercise, remember to switch back into closed-loop mode to regain the benefits of full automation, especially overnight.
Important: this kind of strategy should always be individualised and ideally discussed with your diabetes care team before implementation.
Final thoughts
Contributing to this consensus statement has been a deeply rewarding experience. The hope is that these tools help people with type 1 diabetes to move more, with greater safety and confidence, and with fewer glucose surprises.
By combining practical tactics (targets, modes, timing, and carbohydrate) with an experimental mindset, it is possible to build a personal exercise framework that works with your AID system rather than against it.
For a more detailed exploration of the underlying data, physiological rationale, and system-specific nuance, I strongly encourage you to read the full consensus document and to discuss specific questions with your own diabetes team.
Ready for more of a conference-level deep dive? Have a look at this exploration of AID and exercise from recent scientific meetings.
References and resources
-
EASD–ISPAD. Automated Insulin Delivery and Exercise Consensus Statement. Accessible via PubMed: https://pubmed.ncbi.nlm.nih.gov/39653802/
-
Supplementary graphics (mmol/L): Download PowerPoint
-
Supplementary graphics (mg/dL): Download PDF
-
GNL starter guide to exercise and type 1 diabetes: https://theglucoseneverlies.com/exercise/
-
Conference deep dive on AID and exercise: https://theglucoseneverlies.com/aid-exercise-type-1-diabetes-key-2/
