Does lifting weights lower blood sugar? Regular resistance training can improve longer-term glucose control in adults with type 2 diabetes, but a hard lifting session can temporarily raise a reading.
Those statements describe different time windows. The useful answer separates a reading during a workout, a later response, and the results of training over weeks or months.
Lifting and blood sugar: four different questions
| Question | Relevant evidence | What it cannot establish |
|---|---|---|
| Can glucose rise during or after hard lifting? | The American Diabetes Association describes a temporary hormone-related rise with intense activity, including heavy weightlifting. | Whether one person’s high reading is harmless or requires action. |
| Can glucose fall after activity? | ADA guidance explains contraction-related glucose uptake and increased insulin sensitivity during and after exercise. | A universal size or duration of the response to a particular lifting session. |
| Can regular lifting improve average glucose control? | Human type 2 diabetes trials measure outcomes such as HbA1c after structured training. | A guaranteed result from any workout or a prediction for a reader without that diagnosis. |
| Does lifting outperform running? | A recent direct comparison used a weightlifting model and voluntary wheel running in male mice. | Which exercise is superior for a person’s diabetes care. |
A reading at one moment, a longer-term HbA1c outcome, and an insulin-tolerance test in a mouse are not interchangeable. A headline can become misleading when it keeps the result but drops the population or time window.
Why can blood sugar rise after lifting weights?
Muscles use glucose during activity, but the body can also release glucose into the blood. The ADA explains that adrenaline during intense activity can stimulate liver glucose release, and it names heavy lifting as one example.
This is why an immediate rise and a useful long-term training response can coexist. It does not mean that every rise has the same cause or that an individual reading should be dismissed.
The ADA’s separate exercise guidance explains that contracting muscles can take up glucose and that insulin sensitivity can increase with activity. A later fall is another possible part of the response, so the direction need not remain constant.
This article does not supply glucose thresholds, carbohydrate treatment, insulin adjustments, or an exercise-clearance rule. Those decisions need the person’s established care plan and clinical context.
What human resistance-training trials show over time
A 2011 systematic review and meta-analysis examined structured exercise training in adults with type 2 diabetes. Its pooled resistance-training result favored lower HbA1c compared with control, with substantial variation among studies.
The resistance-training estimate was 0.57 percentage points lower, with a 95 percent confidence interval from 0.01 to 1.14 percentage points lower. That wide interval and the differences between studies make it unsuitable as a promised change for an individual.
A useful counterexample is the nine-month HART-D randomized trial of 262 adults with type 2 diabetes. Its combined aerobic-and-resistance group improved HbA1c compared with control, while the resistance-only and aerobic-only comparisons did not reach statistical significance in the main analysis.
The individual trial and the pooled review are not the same experiment. Their results show why “lifting can help on average” should not become “every lifting plan must lower every person’s HbA1c.”
Nor does a nonsignificant result prove that an intervention has exactly zero effect. It tells us that the trial did not establish a clear difference for that comparison under its own conditions.
Why the lifting-versus-running headline needs a species label
The Virginia Tech-led study compared exercise models during eight weeks of high-fat feeding in male mice. The resistance model involved lifting a weighted food-access lid, while the endurance model used voluntary wheel running.
The resistance group had stronger results on several glucose- and insulin-related outcomes in that design. The authors also reported that the advantage was not explained by significant differences in muscle-weight gain or exercise performance.
That is a preclinical finding worth studying, not a human exercise prescription. The paper did not assign adults with diabetes to ordinary gym lifting and running programs.
It therefore cannot establish that someone should replace walking, running, or other aerobic activity with weights. Human exercise decisions need human evidence plus the individual’s goals and clinical circumstances.
How to read a blood-sugar claim without mixing up the numbers
The NIDDK explains that an A1C test reflects average blood glucose over about three months. A meter reading at one moment is a different measure, even when both appear in an article about blood sugar.
| What you see | Common mistake | More accurate reading |
|---|---|---|
| A higher reading after one workout | “Lifting cannot help glucose control.” | One response cannot answer the longer-term training question. |
| A pooled HbA1c improvement | “My next test will fall by exactly that amount.” | A study average and its uncertainty describe the evidence, not a personal forecast. |
| A resistance advantage in mice | “Weights beat cardio for everyone with diabetes.” | The result belongs to the animal model and its particular outcomes. |
Also check whether an A1C change is expressed in percentage points or as a relative percentage. In a purely fictional arithmetic example, 8.0 percent to 7.5 percent is a 0.5 percentage-point fall, not a 0.5 percent relative fall.
Those fictional values are not targets or an example of what training should achieve. They show why copying a number without its units can change its meaning.
Where an evidence answer stops
The ADA notes that people using insulin or medicines that stimulate insulin release can have an exercise-related risk of low glucose. Its guidance directs personal planning to the diabetes care team.
Keep your established care instructions when discussing a new workout or an unusual response. An evidence table cannot determine medication changes, whether a reading requires treatment, or whether a person needs glucose testing in the first place.
If you already keep exercise and glucose records as part of your care, a useful discussion identifies the activity, the timing of the reading, and any relevant change your team has asked you to record. Ask how those observations should be interpreted alongside longer-term results.
The studies here also do not establish that a person without a clinical indication should start sensor or finger-stick monitoring solely because they lift weights. Testing decisions belong with a qualified clinician.
The answer depends on what you mean by lower
Regular resistance training can support longer-term glucose control in people with type 2 diabetes, while a hard session may raise glucose temporarily. Neither a single reading nor an animal comparison settles the whole question.
Use the timing, population, outcome and uncertainty to interpret a claim. Keep personal treatment and exercise decisions with the professionals who know your situation.
Get the next evidence check.
Five practical fitness stories worth your time, every Friday.
One email a week. No product pitches. Unsubscribe in one click.
Sources and limits
- American Diabetes Association: Why Does Exercise Sometimes Raise Blood Sugar?
- American Diabetes Association: Blood Glucose and Exercise.
- NIDDK: The A1C Test and Diabetes.
- Umpierre and colleagues, 2011: structured exercise and HbA1c systematic review.
- Church and colleagues, 2010: HART-D randomized trial.
- Shute and colleagues: weightlifting and wheel running in mice.
Health disclaimer: This article explains exercise research and is not medical advice or an individual diabetes-management plan. If diabetes, another medical condition, or an injury affects your exercise, discuss changes with your care team.
