For chronic low back pain, three or four exercise sessions a week of about 25 to 30 minutes helped as much as doing more, according to a review of 239 randomized trials published September 24 in the British Journal of Sports Medicine. Extra exercise beyond that added little benefit.
If your back has hurt for months, you do not need a daily gym habit to get the benefit these trials measured. A modest routine, kept up for two to three months, matched the larger doses. What this does not cover is pain that shoots down your leg. Only two of the 239 trials included people with that kind of nerve pain, so it needs its own conversation with your doctor.
Exercise is one of several options people try for a stiff, sore back. A simple lumbar belt also beat usual care in one recent trial, which raises the question of what works best alongside a workout routine.
Key takeaways
- Across 239 trials, low to moderate exercise doses improved back pain and disability in the short term, and higher doses added little.
- The best modeled doses were 3 to 4 sessions a week of 25 to 30 minutes, for 8 to 13 weeks.
- The evidence is weak: 84% of trials had a high risk of bias, and almost none studied pain that runs down the leg.
What the study found
Researchers from Germany, Australia, the UK and elsewhere pooled randomized trials of exercise in adults with chronic low back pain, searching the literature through June 3, 2025. They used a dose-response model to estimate how many sessions, how long each one, and how many weeks gave the most benefit.
- For disability (how much back pain limits daily life), the model’s best dose in the short term was 4 sessions a week, 25 minutes each, over 13 weeks.
- For pain intensity, the best short-term dose was 3 sessions a week, 30 minutes each, over 8 weeks.
- The researchers set a change of at least 5 points on a 0 to 100 scale as the smallest change that matters to a patient.
- Few dose combinations reached that meaningful level beyond the short term, which the review defined as up to 3 months.
The “best” doses are not sharply better than their neighbors. The researchers reported that the credible intervals of most dose levels overlapped. In plain terms, the model cannot confidently separate one dose from the next, so the numbers are a guide rather than a precise target.
Nearly all trials (99.2%) studied non-specific chronic low back pain, meaning back pain without a clear structural cause. Adverse events were reported in 28.5% of trials and were mostly minor muscle and joint complaints.
Dr. Kumar’s take
“More isn’t better” is freeing news. Many people with a bad back assume they have failed unless they exercise every day. This review says a few half-hour sessions a week carry most of the measurable benefit.
The second point matters just as much. Sciatica and other nerve symptoms, where pain travels down the buttock and leg, were almost absent from these trials. The authors themselves call exercise for these radicular presentations a key evidence gap. Nerve root pain has different causes and different warning signs, so these dose numbers should not be applied to it.
Third, the benefits were mostly short term. Few doses held a meaningful effect past three months.
Finally, the evidence quality is poor. With 84% of trials at high risk of bias and certainty rated low to very low, these doses are a reasonable starting point, not settled science.
What it means for you
- If you have chronic low back pain without leg symptoms: about three or four 25 to 30 minute sessions a week, for two to three months, fits what these trials found. Doing more is fine if you enjoy it, but it is not required.
- Keep going after the first few months. The gains faded with time in most doses studied.
- If your pain runs down your leg, or comes with numbness or weakness, this review does not apply to you. Talk with your doctor about a plan built for nerve pain.
- New or worsening back pain should be checked by your doctor before you start a program.
For people weighing pills against movement, whether opioids help much with short-term pain has its own evidence worth knowing.