Exercise and Activity in Type 1 Diabetes:
Why Glucose Moves When You Move
Everything in one place. Read the plain version with Jude, earn your way into the evidence with Grace, then the full model with John. Stop wherever you have enough.
How we teach: three rules, borrowed from Taleb
You earn each level by showing you understand it, not by scrolling past it. We only teach what we would use on ourselves and the people we love.
Understanding beats memory and luck, so the checks reshuffle every time you retry. A pass means you got it, not that you guessed it. And we teach you to tell a trend (signal) from one reading (noise).
We give you the scaffolding and get out of your way. Roam where your curiosity leads, go as deep as you want, and ask Grace anything. We will not teach a bird how to fly.
Want this for your own routine, in your units? Ask Grace, then take it to your care team.
One page, three depths
This guide compounds: each layer rests on the one beneath it. Read Jude’s plain version, then pass a short understanding check to open Grace, then another to open John. You can roam freely within a layer; you cannot skip ahead a layer, because the next one would not make sense and you would be standing on a gap.
The whole thing, in plain words
When you move, your muscles burn glucose, and they can do it whether or not insulin is around. That is why activity is one of the most reliable ways to bring a high down, and also why it can take you low if there is a lot of insulin still working from a recent meal or correction. So the first question before you exercise is not “what shall I do?” It is “how much insulin is still on board, and which way is my glucose heading right now?”
The kind of exercise matters too, and it points in different directions. Steady, breathing-harder activity like a brisk walk, a jog, a swim, or a bike ride usually pulls glucose down. Short, hard, lift-and-strain activity like sprints, heavy weights, or a flat-out effort can push glucose up for a while, because the body releases its own glucose-raising hormones under that kind of stress. Most real sport is a mix of both, so most days land somewhere in between.
Here is the part people miss. The drop does not stop when you do. Muscles keep topping up their glucose stores for hours afterwards, so a low can arrive that evening or overnight, long after the session. This is normal and plannable, not a failure. None of this is a rule to follow alone: keep fast-acting carbohydrate with you, and agree your own plan with your diabetes team, because how much you move, how much insulin you take, and what you eat around it all fit together differently for every person.
Does this match the life of the person living it? Nobody checks a chart mid-football match. The aim is a plan simple enough to actually use, a glucose you are happy to start on, and fast carbs in your pocket, so movement stays a thing you do, not a thing you fear.The Pemberton lens, lived recognisability, one of the four GNL appraisal lenses.
The numbers underneath
Exercise type sets the direction
Across the literature the pattern is consistent: aerobic exercise tends to lower glucose, anaerobic and resistance work often raises it, and mixed sessions sit in between.1 Aerobic work (continuous, oxygen-fuelled effort) drives the largest fall per unit time because contracting muscle keeps pulling glucose in. Anaerobic work (sprints, heavy lifting) triggers a surge of counter-regulatory hormones (catecholamines), which can briefly raise glucose.2 A running session typically drops glucose more than the same time cycling, because more muscle is working (Eckstein 2023).3
| Exercise type | Typical glucose direction | Main driver |
|---|---|---|
| Aerobic (jog, swim, cycle) | Down, often steeply | Sustained muscle glucose uptake |
| Mixed (team sport, circuits) | Variable, often net down | Both effects competing |
| Anaerobic (sprint, lift, HIIT) | Often up, then down later | Counter-regulatory hormones |
Insulin on board is the lever, not the exercise
In a real-world study of close to 500 adults logging nearly 10,000 sessions (the T1DEXI registry), the factors that best predicted an exercise low were, in order: the starting glucose, the CGM trend arrow, the amount of insulin on board, and only then the type of exercise (Bergford 2023).1 The lesson is that exercise hypos are usually caused by supercharged insulin action, not by exercise itself. Within roughly 90 minutes of a meal bolus, insulin is near its peak, which is why a reduced bolus is so often needed before planned activity.
The foundational adjustment scale (Rabasa-Lhoret 2001) reduces the meal bolus before exercise by roughly 25, 50 or 75 percent depending on how long and how hard the session is.4 These are population-average starting points to discuss with your team, not doses to apply unsupervised.
The drop that arrives later
After exercise, muscles refill their glucose stores, and insulin sensitivity stays raised for hours. This is why a delayed and overnight low is common after daytime activity. One well-cited measure is a 0.4 g/kg bedtime carbohydrate snack to reduce overnight hypoglycaemia after an afternoon session (Campbell 2015).5 On time in range, the larger picture is reassuring: in the T1DEXI dataset, exercise days carried lower overall glycaemic risk than sedentary days, with time-above-range falling by more than time-below-range rose (Morrison 2024).6
Schematic shapes, not your data: aerobic effort (teal) lowers glucose during the session; resistance and sprint effort (purple) can raise it first. Both leave raised insulin sensitivity, so a delayed low can follow hours later.
“Exercise raises hypo risk” sounds alarming until you ask the honest questions: which exercise, how much insulin was on board, and over what window? Most of the danger lives in the insulin and the timing, not the movement. Separate the real signal from the scary headline.The Goldacre lens, evidence-grade discipline, one of the four GNL appraisal lenses.
The physiology, the AID era, and the open questions
The mechanism, and why T1D is different
Exercise lowers glucose chiefly through contraction-stimulated GLUT4 translocation, a muscle-uptake pathway that is largely insulin-independent.7 That is why activity works even when insulin is scarce. The catch specific to type 1 diabetes is the broken safety net: the alpha-cell glucagon counter-regulatory response is blunted, so the automatic glucose rescue a non-diabetic body mounts is largely absent.8 The defence has to be exogenous: insulin reduction, carbohydrate timing, and trend-aware decisions. One neat exploitation of physiology is the sprint: a single 10-second maximal effort can release enough catecholamine to blunt the post-exercise fall without extra carbohydrate (Bussau 2006).9
High glucose with raised ketones means too little insulin, and exercise can make ketoacidosis worse, not better. GNL’s conservative reading of the ISPAD chapter is that pump and AID users should not exercise at ketones at or above 0.6 mmol/L, and injection users at or above 1.5 mmol/L; treat the high and recheck first.10 This is one place where there is a right answer, set with your care team.
Exercise on an automated insulin delivery system
An AID system responds to glucose, not to effort. It can suspend or reduce basal, but it cannot recall a bolus already given; that is the insulin-on-board trade-off behind every AID-plus-exercise decision. The practical consequence is that planned activity is an announcement problem, not a reactive one: raising a temporary target or enabling exercise mode 60 to 90 minutes before a session reduces during-exercise hypoglycaemia across systems.11 The four UK systems each have their own control.
| System | Exercise control | Lead time before | Raised target |
|---|---|---|---|
| MiniMed 780G | Temp Target | 60 to 90 min | 8.3 mmol/L (150 mg/dL) |
| Tandem Control-IQ | Exercise Activity | 30 to 60 min | 7.8 to 8.9 mmol/L (140 to 160) |
| CamAPS FX | Ease-off mode | ~60 min | user-selectable, up to 11 mmol/L (200 mg/dL) |
| Omnipod 5 | Activity Feature | can be immediate | 8.3 mmol/L (150 mg/dL) |
A The announce-and-raise-target principle is consensus-backed (EASD/ISPAD 2025).11 D The per-system lead times and target values are a GNL working summary of system documentation and sparse head-to-head data; treat exact numbers as starting points to confirm with your team. Critical rule: never confuse your day-to-day glucose target with the exercise-mode target; they are different controls.
What we still do not know
The honest edges of the evidence: head-to-head comparisons of how the four AID systems handle the same exercise protocol are sparse; the optimal pre-exercise lead time (60 vs 90 vs 120 minutes) is not well quantified; anaerobic handling is less evidenced than aerobic, because an algorithm cannot anticipate a hormone-driven spike; and the post-exercise window data is dominated by adults, with paediatric and sex-difference signals still emerging (Wang 2024; Yardley 2023).1 Where GNL returns carbohydrate figures, the per-kilogram dose is weight-driven but capped at 60 kg, a conservative ceiling so a heavier body is not handed an unsafely large carbohydrate suggestion; insulin-per-kilogram is left uncapped.
It is the rare, large event that does the damage: the unplanned low mid-session, or the one overnight you sleep through. Plan for the fat tail. Carry more fast carbohydrate than an average day needs, keep a starting glucose with margin, and never optimise the average Tuesday at the cost of the catastrophic low you cannot afford.The Taleb lens, robustness to outliers, one of the four GNL appraisal lenses.
A model is only as honest as its assumptions. The exercise framework rests on a handful of small mechanistic studies and one large real-world registry, leaning hard on adult data and a single consensus statement. That is a clean lens, not the whole view. Name what would strengthen it (more paediatric, more head-to-head AID data), and never sell the framework as the territory.The Hayes lens, technical and methodological rigour, one of the four GNL appraisal lenses.
The whole guide, summarised
Glucose never lies; it just reacts honestly to a fast, powerful physiology. Read the lever before you read the exercise, and plan for the hours after.
This page is the taster. The full journey, three modules and their 30 questions, with your progress saved, lives in Learn with Grace. Glucose never lies; come and learn to read it before you move.
References
Evidence grades A (strongest) to D (editorial or working analysis).
- Bergford S, et al. Factors associated with hypoglycaemia during exercise in adults with type 1 diabetes (the T1DEXI real-world registry, close to 500 adults, nearly 10,000 sessions). 2023. Predictor ordering: starting glucose, CGM trend, insulin on board, exercise type. A
- Riddell MC, et al. Exercise management in type 1 diabetes: a consensus statement. Lancet Diabetes Endocrinol. 2017;5(5):377-390. (Aerobic lowers, anaerobic raises; counter-regulatory mechanism.) A
- Eckstein ML, et al. Glucose responses to running versus cycling in type 1 diabetes. Diabet Med. 2023. (Running drops glucose more, greater muscle mass engaged.) A
- Rabasa-Lhoret R, et al. Guidelines for premeal insulin dose reduction for postprandial exercise of different intensities and durations in type 1 diabetes treated intensively. Diabetes Care. 2001;24(4):625-630. (The 25 / 50 / 75 percent bolus reduction scale.) A
- Campbell MD, et al. Bedtime carbohydrate to prevent overnight post-exercise hypoglycaemia (0.4 g/kg). J Clin Endocrinol Metab. 2015. A
- Morrison D, et al. Glycemia Risk Index on exercise versus sedentary days, a post-hoc T1DEXI analysis (n=408 adults). J Diabetes Sci Technol. 2024. DOI 10.1177/19322968241246458. (Exercise days lower net glycaemic risk; AID floor effect.) A
- Sylow L, et al. Exercise-stimulated glucose uptake, regulation and implications for glycaemic control. (Contraction-driven, largely insulin-independent GLUT4 translocation.) Cell Metabolism / review. A
- Colberg SR, et al. Glucagon and counter-regulation during exercise in type 1 diabetes (the blunted alpha-cell response). 2022 review. A
- Bussau VA, et al. The 10-second maximal sprint, a novel approach to counter an exercise-mediated fall in glycaemia in individuals with type 1 diabetes. Diabetologia. 2006. A
- Adolfsson P, et al. ISPAD Clinical Practice Consensus Guidelines 2022, Chapter 14: Exercise in children and adolescents with diabetes. (Ketone safety ladder; GNL applies a conservative Grade-D synthesis of the pump and AID band.) A
- Moser O, Zaharieva DP, Pemberton J, et al. Position statement on exercise and automated insulin delivery in type 1 diabetes (EASD / ISPAD 2025). Diabetologia. (Announce activity, raise temporary target, per-system exercise modes.) A
One page, three voices: Jude, Grace, John. Population-average, not personalised.