Antibiotics for Modic changes rest on a hypothesis: that a low-grade infection drives some cases. Two placebo-controlled trials in people with a previous disc herniation disagree. Albert 2013 reported a large benefit. The larger AIM 2019 trial found less than a clinically important difference. A 2026 Cochrane review has low confidence in small benefits. Albert · 2013 Bråten · 2019 Liu · 2026
What is the hypothesis?
One proposal is that a low-grade infection involving bacteria such as Cutibacterium acnes can contribute to disc and endplate changes after a herniation. Alternative explanations and contamination of samples complicate that argument. The MRI pattern cannot demonstrate that an infection exists. Liu · 2026
The first trial
Albert 2013: a large reported benefit
Albert · 2013Median RMDQ disability fell from 15 to 7 with antibiotics and from 15 to 14 with placebo at one year.
- Who was studied
- 162 adults with chronic back pain after disc herniation and Type 1 change
- Compared with
- Antibiotic treatment versus placebo
- Outcome & follow-up
- RMDQ disability, 0–24 scale · 1 year
- Limits
- Later AIM results conflicted; selected population and trial limitations matter.
- Funding
- Foundation funding reported; author treatment interests discussed in publication.
The larger replication
AIM 2019: below the clinically important difference
Bråten · 2019At one year, antibiotics improved disability by 1.6 RMDQ points more than placebo, below the trial’s pre-specified 4-point clinically important difference.
- Who was studied
- 180 adults with chronic back pain, prior disc herniation and Modic 1 or 2
- Compared with
- Amoxicillin versus placebo
- Outcome & follow-up
- RMDQ disability, 0–24 scale · 1 year
- Limits
- Type 1 subgroup difference 2.3 points; subgroup analysis does not prove a general treatment effect.
- Funding
- Norwegian public health authorities Helse Sør-Øst and Helse Vest.
Adverse events were more common with antibiotics
Bråten · 2019Drug-related adverse events occurred in 56% of the antibiotic group and 34% of the placebo group.
- Who was studied
- 180 adults with chronic back pain, prior disc herniation and Modic 1 or 2
- Compared with
- Amoxicillin versus placebo
- Outcome & follow-up
- At least one drug-related adverse event · During the trial
- Limits
- This is one regimen in a selected trial; it does not quantify every antibiotic risk.
- Funding
- Norwegian public health authorities Helse Sør-Øst and Helse Vest.
What the review concluded
Cochrane 2026: low confidence in benefit
Liu · 2026Antibiotics may give small benefits for back pain and small to moderate benefits for disability, with low confidence in the evidence.
- Who was studied
- People with Modic Type 1 change and evidence of disc herniation
- Compared with
- Amoxicillin with or without clavulanate versus placebo
- Outcome & follow-up
- Pain and back-related disability · 12–14 weeks
- Limits
- Evidence on adverse events is very uncertain; studies were limited and results varied.
- Funding
- No dedicated funding for the review.
The review is by Liu and colleagues, published on 7 April 2026, with evidence searched to August 2025. It is a review of treatment evidence, not a recommendation to self-treat. Liu · 2026
Why the trial population matters
A previous disc herniation at the relevant level was central to the original trials. Results cannot simply be applied to a person whose only finding is “Modic change”. Type 1 and Type 2 subgroup results should also not be treated as a personalised forecast. Bråten · 2019
Harms and alternatives
Long antibiotic courses can cause gastrointestinal symptoms and allergy, and contribute to antimicrobial resistance. Discuss any suspected infection with a clinician. This guide gives no antibiotic dose or treatment course to follow. Conservative care and investigation of other pain sources remain part of that conversation.
Six questions to take with you.
You do not need to understand every technical term before your appointment.
- Where are the changes, and which MRI sequences show them?
- What makes you think they do, or do not, explain my pain?
- What other possible pain sources have you considered?
- What is a realistic plan for conservative care, and how will we review progress?
- For any proposed treatment, how closely do I match the people in its trials?
- What are the benefits, harms, alternatives, costs and reasons it might not help?