Start Where You Are, Not Where You Think You Should Be
The most common mistake I see patients make when they decide to start exercising is going too hard, too fast, and burning out within weeks. Someone who hasn't exercised in years reads that high-intensity interval training (HIIT) produces the greatest A1c reductions and decides that's what they need to do — then quits after two sessions because it's painful and exhausting. The data on HIIT efficacy is real. The data on long-term HIIT adherence in sedentary adults is far less impressive.
The single most important principle in exercise for diabetes management is sustainability. A moderate walking routine sustained for years produces more benefit than an aggressive gym program abandoned after a month. Every piece of exercise advice that follows should be read through that lens: what can this person actually do, consistently, given their current fitness level, their schedule, their physical limitations, and their life?
When I talk to a newly diagnosed patient about exercise, I don't start with what's optimal — I start with what's feasible. If someone is currently sedentary, I ask what they like, what they can tolerate without pain, and when they can realistically fit movement into their day. Walking is almost always the starting point. It requires no equipment, no gym membership, no special skill, and the evidence for its benefits in diabetes is solid. We build from there. The goal is a sustainable habit, not a heroic performance.
What Exercise Actually Does to Blood Sugar
Understanding the mechanism makes the prescription make more sense. During exercise, skeletal muscle contracts and takes up glucose from the bloodstream — and crucially, this uptake happens through a pathway that does not require insulin. A specific glucose transporter called GLUT4 moves to the surface of muscle cells in response to muscle contraction, independently of insulin signaling. This is why exercise lowers blood glucose even in people with severe insulin resistance or very limited insulin production: it bypasses the broken lock entirely.
The effect on insulin sensitivity persists well beyond the exercise session itself. A single bout of moderate aerobic exercise improves insulin sensitivity for 24–72 hours afterward, and regular training produces structural adaptations — more mitochondria in muscle cells, greater GLUT4 density, improved cardiovascular efficiency — that compound over months. This is not the same as a medication wearing off. The benefits of regular exercise accumulate and persist.
There is an important nuance for Type 1 diabetes and for high-intensity exercise: intense anaerobic activity (sprinting, heavy resistance training, HIIT) can temporarily raise blood glucose rather than lower it, because the stress hormones released during intense effort — adrenaline and cortisol — signal the liver to release glucose and counteract insulin. This is predictable and manageable, but it means that the glucose response to exercise depends on the type, intensity, and duration of the activity, and patients on insulin need to understand that different exercise sessions may require different adjustments. We cover this in more detail in Post 7 — Monitoring and Insulin Management.
The Evidence: What Exercise Achieves in Type 2 Diabetes
The evidence base for exercise in Type 2 diabetes is extensive and consistent. The headline numbers are meaningful by any clinical standard.
A1c reduction: Exercise training reduces A1c by 0.6–0.8 percentage points in meta-analysis — comparable to the effect of metformin in the UKPDS trial. A dose-response meta-analysis found each 30 minutes per week of supervised aerobic training reduces A1c by approximately 0.22 percentage points, with maximum benefit at around 100 minutes per week (−0.96% reduction), after which the curve flattens.
Cardiovascular mortality: High versus low physical activity is associated with 29% lower cardiovascular mortality (RR 0.71) and 39% lower all-cause mortality (RR 0.61) in people with diabetes. Each additional 1 MET-hour per day of physical activity is associated with a 9.5% reduction in all-cause mortality and 7.9% reduction in cardiovascular disease risk.
Weekend warrior data (2025 cohort, 51,650 US adults with diabetes): Meeting physical activity recommendations of ≥150 minutes per week was associated with 21% lower all-cause mortality — whether that activity was spread through the week (HR 0.83) or concentrated on weekends (HR 0.79 for "weekend warriors"). The weekend warrior pattern was associated with 33% lower cardiovascular mortality (HR 0.67). Activity pattern matters far less than total volume.
Microvascular complications: High versus low physical activity is associated with 24% lower risk of major diabetes-related complications including kidney disease and neuropathy (RR 0.76, moderate certainty evidence).
Cognitive function: Meta-analysis of 12 RCTs found exercise improved cognitive function with a standardized mean difference of 0.91 — a meaningful effect in a population at elevated dementia risk. Multimodal exercise (combining aerobic and resistance) produced the greatest cognitive benefit.
The Weekend Warrior Finding Deserves Emphasis
I highlight the weekend warrior data specifically because it removes one of the most common barriers patients cite: "I can't exercise during the week — my schedule doesn't allow it." The 2025 cohort study of over 51,000 adults with diabetes found that concentrating 150 minutes of moderate-to-vigorous activity on weekends produced cardiovascular mortality reductions that were actually slightly larger than those seen in people who spread activity evenly throughout the week.
This does not mean weekday activity is unimportant — more activity is better than less, and spreading it across the week has benefits for daily glucose management that weekend-only exercise cannot fully replicate. But for patients who cannot currently achieve daily movement, two longer weekend sessions represent a legitimate, evidence-backed starting point that is meaningfully better than waiting for a perfect schedule to materialize.
51,650 US adults with diabetes followed for mortality outcomes. Compared to inactive participants:
- Regularly active (≥150 min/week spread across multiple days): All-cause mortality HR 0.83, cardiovascular mortality HR 0.81
- Weekend warriors (≥150 min/week mostly on 1–2 days): All-cause mortality HR 0.79, cardiovascular mortality HR 0.67
- Both patterns significantly superior to inactivity. Weekend warrior cardiovascular benefit was larger — likely reflecting that those who exercise intensely on weekends tend to accumulate higher total activity volumes.
Exercise in Type 1 Diabetes: More Complex, Still Essential
The evidence for A1c improvement with exercise in Type 1 diabetes is less clear-cut than in Type 2 — not because exercise is less beneficial, but because the picture is more complicated. In Type 1, the relationship between exercise and blood glucose depends heavily on insulin on board, timing relative to meals, exercise type, and individual response. Managing these variables is genuinely difficult, and some patients experience glucose instability that makes exercise feel more hazardous than helpful.
However, the benefits in Type 1 are well-established across other domains. Exercise is associated with improved cardiovascular health, better lipid profiles, improved insulin sensitivity (and correspondingly reduced insulin requirements), reduced risk of microvascular complications, and better bone density. A large real-world dataset of over 428,000 exercise sessions in people with Type 1 found that all exercise modes — aerobic, high-intensity interval, and resistance — improved time-in-range over the 24 hours following the session, with anaerobic exercise producing the largest improvements. The glucose management complexity is real but navigable, particularly with continuous glucose monitoring (CGM).
A 2026 analysis of 428,058 exercise sessions in people with Type 1 diabetes found:
- All exercise modes improved time-in-range (the percentage of time glucose stayed between 70–180 mg/dL) in the 24 hours following exercise
- Aerobic exercise acutely lowers blood glucose — requiring pre-exercise carbohydrate or insulin adjustment
- Anaerobic exercise (resistance training, HIIT) produced the largest 24-hour time-in-range improvements, partly because the acute glucose-raising effect of stress hormones is followed by sustained improved insulin sensitivity
- Meeting step count targets of ~7,000–10,000 steps per day was associated with marginal improvements in time-in-range and reduced insulin requirements
For patients with Type 1, I often recommend starting with resistance training rather than aerobic exercise, for a practical reason: resistance exercise raises glucose acutely (through stress hormone release), which means hypoglycemia during the session is less likely. Aerobic exercise lowers glucose during activity — useful when glucose is high, but a management challenge when insulin is on board. Using CGM to observe personal glucose responses to different exercise types is one of the most useful tools available for building a Type 1 exercise routine. What works for one person won't be identical for another, and the only way to know is to watch the data.
Types of Exercise: What Each One Does
Resistance Training
Resistance training — weights, resistance bands, bodyweight exercises — builds and preserves skeletal muscle, which is the body's primary glucose-disposal tissue. More muscle means more capacity to absorb glucose from the bloodstream after meals, independent of how much insulin the pancreas is producing. This is why resistance training is particularly valuable in Type 2 diabetes and becomes more important with age, as muscle mass naturally declines.
Meta-analysis shows A1c reductions of approximately −0.57% with resistance training alone, comparable to aerobic exercise. In Type 1 diabetes, resistance training raises blood glucose acutely due to stress hormone release, which makes it lower-risk for hypoglycemia during the session — an important practical advantage. The ADA recommends 2–3 sessions per week on non-consecutive days.
Aerobic Exercise
Aerobic exercise — walking, cycling, swimming, dancing, any sustained activity that elevates the heart rate — is the most studied form of exercise in diabetes and has among the best cardiovascular evidence of any intervention. It improves heart efficiency, lowers blood pressure, reduces triglycerides, and directly lowers blood glucose during and after the session by driving GLUT4-mediated glucose uptake in muscle.
The dose-response curve is meaningful: maximum A1c benefit in meta-analysis occurs around 100 minutes per week, with each 30-minute weekly increment reducing A1c by approximately 0.22 percentage points up to that point. Walking specifically has been shown safe even in patients with peripheral neuropathy and does not increase foot ulcer risk in the absence of active foot disease. For Type 1 patients, aerobic exercise typically requires carbohydrate supplementation or insulin reduction to prevent hypoglycemia during activity.
Combined Aerobic + Resistance
Combining aerobic and resistance training in the same week produces benefits across the broadest range of outcomes — glucose control, cardiovascular fitness, muscle mass, bone density, and cognitive function. The ADA recommends this combined approach as the most complete exercise prescription for adults with diabetes.
In practice, this might look like two resistance sessions and three walking sessions per week — not an overwhelming commitment, but producing benefits that neither alone fully achieves. For patients who can only do one type, either is valuable; the combination is the goal when feasible.
High-Intensity Interval Training (HIIT)
HIIT — alternating brief bursts of near-maximal effort with recovery periods — produces the largest A1c reductions in meta-analysis (−0.78%), likely by driving greater GLUT4 expression and greater metabolic stress adaptation than steady-state exercise. It also requires less total time than continuous moderate exercise to achieve comparable glucose benefits.
The problem is that HIIT is genuinely hard, and hard exercise has poor long-term adherence in people who are not already fit and exercise-motivated. I don't start patients on HIIT. If a patient is already reasonably fit and wants to maximize metabolic benefit with limited time, HIIT is a legitimate recommendation. For someone who is deconditioned, in pain with exertion, or has never exercised consistently, starting with HIIT is a reliable way to ensure they stop exercising. Build the base first.
Yoga and Mind-Body Exercise
Yoga produces meaningful A1c reductions (approximately −0.45%) and has the added benefit of reducing cortisol and stress-related glucose elevation — a pathway that aerobic exercise doesn't address as directly. Chronic stress raises blood glucose through cortisol-mediated insulin resistance, and this is an underappreciated driver of poor glucose control in some patients.
Yoga and tai chi are also particularly useful for older adults or those with limited mobility or balance concerns, where higher-impact exercise is not appropriate. They're not a replacement for aerobic or resistance training when those are possible, but they are a legitimate addition — and for some patients, the entry point that makes regular movement sustainable.
Exercising Safely With Diabetes Complications
Most patients with diabetes can and should exercise. But specific complications require specific modifications. These are not reasons to avoid exercise — they are reasons to exercise thoughtfully.
Medication Interactions and Hypoglycemia Risk
Not all diabetes medications carry the same hypoglycemia risk during exercise. This matters because hypoglycemia during exercise is the most common reason patients with diabetes avoid being physically active — and for some drug combinations, that fear is well-founded. For others, it is largely unnecessary. A full discussion of each medication class is in Post 3 — Treatment; what follows here covers the exercise-specific interactions.
Pre-exercise glucose targets for Type 1
The ADA recommends a target blood glucose of 90–250 mg/dL before starting exercise in Type 1 diabetes. Below 90 mg/dL, take 10–15g of fast-acting carbohydrates before beginning low-to-moderate intensity exercise. Above 250 mg/dL, check for ketones — if significant ketones are present, exercise should wait until glucose and ketones are controlled. Glucose above 300 mg/dL without ketones allows cautious exercise with close monitoring and good hydration. Post-exercise hypoglycemia — occurring several hours after activity — is common after longer or more intense sessions and requires monitoring and sometimes a bedtime snack if the session was in the evening.
The ADA Recommendation — and How to Get There
The ADA recommends at least 150 minutes per week of moderate-to-vigorous aerobic activity spread over at least 3 days, with no more than 2 consecutive days without activity, plus 2–3 resistance training sessions per week on non-consecutive days. For someone currently sedentary, this can feel like an overwhelming target.
The practical path: start with whatever movement is sustainable without pain. For most people that is 10–15 minutes of walking. Build gradually — adding 5 minutes per session every 1–2 weeks. Add resistance training once aerobic activity is established and comfortable. The 150-minute target is a goal to work toward over weeks to months, not a minimum requirement for starting. Any movement is better than no movement. The evidence shows clear, graded benefit even at modest activity levels.
The patients I see make lasting changes with exercise are the ones who find something they don't hate, that fits their actual schedule, and that they can do even on hard days. That might be a 20-minute walk at lunch, a Saturday morning bike ride, or a resistance band routine in front of the television. The form matters far less than the consistency. I'd rather have a patient walking 30 minutes most days than doing a medically ideal program for three weeks before quitting. Build the habit first. Optimize later.