Prednisone for Back Pain: Not Recommended
Systemic corticosteroids, including prednisone, are not recommended for the treatment of back pain, as high-quality evidence consistently demonstrates no clinically meaningful benefit for pain relief or functional improvement in non-radicular low back pain, and only minimal short-term benefits in radicular pain that do not justify the risks. 1
Evidence for Non-Radicular (Mechanical) Back Pain
For acute non-radicular low back pain—the typical mechanical back pain from bending, twisting, or muscle strain—prednisone provides no benefit:
- Two randomized trials found no differences between systemic corticosteroids (either a single intramuscular injection or a 5-day oral course) and placebo for pain or function 1
- A well-designed emergency department trial of 50 mg prednisone daily for 5 days showed no improvement in pain scores, functional status, return to work, or days lost from work compared to placebo 2
- Patients receiving prednisone were actually more likely to seek additional medical treatment (40% vs 18%) than those receiving placebo 2
- For chronic non-radicular back pain, no trials have evaluated systemic corticosteroids 1
Bottom line: Do not prescribe prednisone for mechanical/non-radicular back pain. 1, 2
Evidence for Radicular Back Pain (Sciatica)
For radicular low back pain with leg symptoms, the evidence shows only marginal benefit:
- Six high-quality trials consistently found no meaningful differences in pain between systemic corticosteroids and placebo 1
- Pain improvement was statistically significant but clinically trivial: only 0.56 points better on a 0-10 scale (95% CI: 0.04 to 1.08) 3
- Short-term functional improvement was minimal and inconsistent across studies 1, 3
- One large trial (n=269) showed a small functional benefit at 52 weeks (7.4-point improvement on the Oswestry Disability Index), but this was the only positive finding among multiple trials 1
- Systemic corticosteroids did not reduce the need for spine surgery 1, 3
Adverse Effects Outweigh Minimal Benefits
Even short courses of prednisone cause significant side effects:
- In the largest trial, oral prednisone (starting at 60 mg/day) increased risk for any adverse event (49% vs 24%; P<0.001), insomnia (26% vs 10%), nervousness (18% vs 8%), and increased appetite (22% vs 10%) 1
- These harms occur even with brief treatment courses 1, 3
Evidence for Spinal Stenosis
For spinal stenosis, prednisone is ineffective:
- One trial (n=61) found no differences between a 3-week course of prednisone and placebo in pain intensity or function through 12 weeks of follow-up 1
- A more recent trial of 10 mg prednisolone daily for 1 week in refractory spinal stenosis showed no significant improvement in pain or disability, only a modest increase in walking distance 4
Clinical Recommendation Algorithm
For any patient presenting with back pain:
Non-radicular (mechanical) back pain: Do NOT prescribe prednisone. Use NSAIDs, acetaminophen, muscle relaxants if needed, and reassurance about natural history 1, 2
Radicular back pain (sciatica): Do NOT routinely prescribe prednisone. The minimal benefit (0.56 points on 0-10 scale) does not justify the adverse effects 1, 3. Consider NSAIDs, neuropathic pain medications (gabapentin, pregabalin), and physical therapy instead 1
Spinal stenosis: Do NOT prescribe prednisone—it is ineffective 1, 4
If considering corticosteroids despite lack of evidence: Epidural corticosteroid injections have been studied more extensively than oral therapy and may be considered in select cases of radicular pain, but this is a separate intervention requiring specialist referral 1
Common Pitfalls to Avoid
- Do not prescribe prednisone reflexively for "severe" back pain—severity does not predict response to corticosteroids 1, 2
- Do not use the "anti-inflammatory" rationale to justify prednisone—NSAIDs are safer and equally or more effective 1
- Do not assume short courses are harmless—even 5-day courses cause significant adverse effects including insomnia, mood changes, and increased appetite 1, 2
- Do not confuse epidural with systemic corticosteroids—the evidence discussed here applies to oral/intramuscular routes only 1