The Best Exercise to Lower Blood Sugar: Cardio vs Weights vs HIIT — What 2025-2026 Science Actually Says
If you have type 2 diabetes or prediabetes, you've probably been told to exercise more. What nobody tells you is which exercise actually works best for blood sugar, and when to do it. That answer has changed in the past 12 months.
Three significant studies published between October 2025 and April 2026 now give us the clearest picture yet. Resistance training outperforms running for improving insulin sensitivity (Shute et al., Journal of Sport and Health Science, Oct 2025). A 10-minute walk taken immediately after a meal cuts blood sugar spikes more than a longer walk done 30 minutes later (Hashimoto et al., Scientific Reports, Jul 2025). And exercise duration, not type or intensity, is the single strongest predictor of HbA1c improvement in newly diagnosed type 2 diabetes (Low et al., Medicine & Science in Sports & Exercise, Apr 2026).
The short version: find something you can do for a long time, add weight training if you can, and walk right after you eat.
Key Takeaways
- Exercise duration outranks type, intensity, and frequency for HbA1c improvement (UBCO MOTIVATE T2D trial, Apr 2026)
- HIIT reduces HbA1c by 0.39% to 0.83% vs no exercise (Poon et al., Diabetes Obesity & Metabolism, 2025)
- Resistance training beats running for visceral fat reduction and insulin sensitivity (Virginia Tech, Oct 2025)
- A 10-minute post-meal walk cuts peak blood glucose from 181.9 to 164.3 mg/dL (Scientific Reports, Jul 2025)
- CGM shows which exercise type and timing works best for your individual glucose response
Does Exercise Actually Lower Blood Sugar?
Yes — and the effect kicks in within 24 hours. The American Diabetes Association confirms exercise increases insulin sensitivity, meaning your cells absorb glucose more effectively both during and long after a session. A meta-analysis of 22 randomised controlled trials found that combined aerobic and resistance training improved both HOMA-IR (a key insulin resistance marker) and fasting blood sugar more than either modality alone (Liu et al., Frontiers in Endocrinology, Nov 2025).
The mechanism differs depending on which exercise you choose. Aerobic exercise pulls glucose directly into working muscles during activity. Resistance training builds muscle tissue that acts as a long-term glucose reservoir, absorbing blood sugar even when you're resting. HIIT does both in a compressed timeframe. All three lower blood sugar. They just do it differently.
Cardio vs Weights: Which Wins for Blood Sugar Control?
For most of the past two decades, aerobic exercise was the default diabetes recommendation. A 2025 study from Virginia Tech is shifting that view. Researchers at the Fralin Biomedical Research Institute found that resistance training outperformed endurance exercise in reducing visceral fat, improving glucose tolerance, and lowering insulin resistance in an obesity and type 2 diabetes model (Shute et al., Journal of Sport and Health Science, Oct 2025). Lead researcher Professor Zhen Yan summed it up: "Weight training has equal, if not better, anti-diabetes benefits."
The reason comes down to muscle. Every kilogram you build acts like a permanent glucose reservoir — absorbing blood sugar through an insulin-independent pathway. That matters because insulin resistance is the core problem in type 2 diabetes. More muscle means more glucose cleared from your blood, around the clock.
[UNIQUE INSIGHT] This doesn't mean you should swap your walks for squats. The Virginia Tech team was explicit: combining both produces the best outcome. The finding doesn't retire cardio; it elevates resistance training to equal standing, which most diabetes exercise guides still haven't done. If you've been skipping the weights, that's the gap worth closing first.
What HIIT Does to Blood Sugar — and When It Makes Sense
HIIT is the most time-efficient route to HbA1c reduction, provided you can sustain it. An umbrella review of systematic reviews and meta-analyses found that HIIT consistently reduced HbA1c in type 2 diabetes patients vs no-exercise controls, with weighted mean differences ranging from -0.39% to -0.83% across six separate systematic reviews (Poon et al., Diabetes Obesity and Metabolism, Feb 2025). Low-volume HIIT, shorter sessions using less total work, reduced HbA1c by 0.70% and fasting glucose by 16.63 mg/dL in overweight and obese type 2 diabetes participants specifically (Lu et al., Frontiers in Endocrinology, Jan 2025).
The practical appeal is time. A 20-minute HIIT session three times a week produces comparable HbA1c improvements to 150 minutes of moderate aerobic exercise in many studies. For people in Southeast Asia and Australia juggling demanding work schedules, that trade-off is worth knowing about. The main barrier to exercise in both regions isn't motivation; it's finding enough hours.
The 2026 Finding That Changes How You Should Think About Exercise
The most counterintuitive result of the past year comes from the MOTIVATE T2D trial at the University of British Columbia Okanagan. Researchers tracked 58 newly diagnosed type 2 diabetes adults over 26 weeks using fitness watches and found that exercise duration per session was the single strongest predictor of blood sugar improvement. It outranked exercise type, intensity, calories burned, and frequency (Low et al., Medicine & Science in Sports & Exercise, Apr 2026).
"We saw a clear signal that exercise duration stood out above all other factors," said Dr. Jonathan Little, the trial's lead researcher. "Find something you like and do it for as long as you can."
[UNIQUE INSIGHT] Most diabetes exercise guides lead with intensity: hit 70% of your max heart rate, work hard enough to break a sweat. The UBCO data flips that for inactive, newly diagnosed patients. How long you move matters more than how hard you push. A 45-minute gentle walk beats a 15-minute run for this group. The practical shift is significant: instead of asking "am I working hard enough?", the better question is "can I keep this going longer?"
Citation capsule: The MOTIVATE T2D trial (58 adults, 26 weeks, fitness watch monitoring) found that exercise duration per session was the strongest predictor of HbA1c improvement in newly diagnosed type 2 diabetes, outranking exercise type, intensity, frequency, and calories burned. Lead researcher Dr. Jonathan Little of UBC Okanagan concluded: "Find something you like and do it for as long as you can" (Low et al., Medicine & Science in Sports & Exercise, Apr 2026).
Does the Timing of Your Workout Affect Blood Sugar?
Yes — and a CGM makes the difference visible. A randomised crossover trial found afternoon HIIT kept mean 24-hour CGM glucose at 6.2 mmol/L in men with type 2 diabetes, while the same morning HIIT protocol raised it to 6.9 mmol/L (Savikj et al., Diabetologia, cited in JCEM, Apr 2025). Skeletal muscle mitochondrial function peaks between 16:00 and 20:00 due to circadian changes in neural excitability, which is why the body clears glucose more efficiently in the afternoon.
Timing matters most right after meals. A 2025 randomised crossover trial found that a 10-minute walk taken immediately after eating reduced peak blood glucose to 164.3 mg/dL, compared to 181.9 mg/dL in participants who rested. That's a 9.7% reduction in glucose peak from 10 minutes of light walking at 3.8 km/h (Hashimoto et al., Scientific Reports, Jul 2025). Critically, this 10-minute walk matched the effect of a 30-minute walk done 30 minutes after eating.
[IMAGE: Editorial AI image — person doing a relaxed post-meal walk outdoors in afternoon light, Southeast Asian or Australian urban setting, CGM sensor visible on upper arm, glucose app on phone showing stable reading]
How CGM Turns Exercise Research Into Personal Data
Population-level research tells you what works on average. A CGM tells you what works for your body, in real time. That gap matters more than most people realise, because glucose responses to exercise vary substantially between individuals.
Some people see a sharp drop during aerobic exercise. Others stay flat. HIIT can spike glucose briefly from adrenaline, then produce a sustained drop two hours later. Resistance training sometimes raises glucose slightly during the session, then delivers a 12-24 hour reduction as insulin sensitivity improves. A fasting reading the next morning won't show you any of this. A CGM captures all of it.
how to read your CGM data → guide to understanding time-in-range and exercise patterns
Three practical ways to use a CGM with exercise:
- Track your post-meal window. Walk 5-10 minutes after your largest meal. Check your CGM reading before and 90 minutes after. Within a week, you'll see whether the walk is blunting your glucose spike, and by how much.
- Compare morning vs afternoon sessions. Run the same workout at 7am one week and 5pm the next. Your CGM's 24-hour average will show which timing is working harder for your glucose, the same measurement the Diabetologia crossover trial used.
- Find your sustainable effort level. The UBCO finding shows duration matters most. CGM data helps you identify which activities you can keep going longest at a comfortable pace, because those are the sessions that will shift your HbA1c over time.
Frequently Asked Questions
Can exercise replace diabetes medication?
Exercise is a cornerstone of non-pharmacological diabetes management, not a replacement for prescribed treatment. The Journal of Clinical Endocrinology & Metabolism classifies physical activity as "a staple of lifestyle management" for both type 1 and type 2 diabetes, supporting glycaemic control alongside medication (Lewis et al., JCEM, Apr 2025). Always discuss any changes to your treatment plan with your GP or endocrinologist.
Why does my blood sugar go up during intense exercise?
High-intensity exercise triggers adrenaline and cortisol release, signalling your liver to dump stored glucose for fuel. The temporary rise is normal and typically followed by a sustained blood sugar reduction over the next several hours as insulin sensitivity improves. Low to moderate exercise such as walking tends to lower blood sugar directly during activity, while high-intensity exercise may raise it briefly before the longer-term benefit appears (Medical News Today, Nov 2025).
How many minutes of exercise per week do I actually need?
The ADA recommends at least 150 minutes of moderate-intensity aerobic activity per week. The UBCO MOTIVATE T2D trial adds a useful nuance: for newly diagnosed type 2 diabetes patients, duration per session matters more than hitting a weekly total. Longer individual sessions outperformed shorter, more frequent ones for HbA1c improvement. If you can only exercise three times a week, make each session as long as you comfortably can (Low et al., Medicine & Science in Sports & Exercise, Apr 2026).
Is walking enough, or do I need weights too?
Walking is effective — the 2025 post-meal walk data confirms that even 10 minutes makes a measurable glucose difference. Adding resistance training provides benefits walking alone can't deliver: greater muscle mass, better baseline insulin sensitivity, and superior visceral fat reduction. The Virginia Tech team's recommendation is to do both when possible (Shute et al., Journal of Sport and Health Science, Oct 2025). Starting from scratch? Walking first is exactly the right call.
