Patient-friendly summary
If you read nothing else
Bottom line
CBT may produce small short-term improvements in pain, disability, fear avoidance, and self-efficacy for chronic low back pain — most effective as an add-on to other therapy — but the evidence is low quality and benefits were not clearly maintained at follow-up.
Moderate evidencePublished
Evidence hierarchy
Study participants
Adults (≥18 years) with chronic low back pain (pain duration ≥3 months), with or without leg pain, across nine countries
Study Summary
This systematic review and meta-analysis pooled 22 articles (20 randomized controlled trials, 3,003 patients) to assess whether cognitive-behavioral therapy (CBT) helps people with chronic low back pain (pain lasting over 3 months). Immediately after treatment, CBT showed small but statistically significant advantages over comparison therapies for disability, pain, fear avoidance, and self-efficacy, though most of this came from low-quality evidence with high statistical heterogeneity. No significant differences between CBT and comparison therapies were found at the 3-, 6-, or 12-month follow-ups, and the pain benefit weakened sharply when two outlier studies were removed. The authors conclude CBT may be beneficial, especially when added on top of other therapies (such as physical therapy), but call for higher-quality trials with longer follow-up to confirm lasting benefit.
Key Findings
| Finding | Detail | Impact |
|---|---|---|
| CBT showed small short-term benefits over comparison therapies for pain and disability | Immediately after intervention, pain SMD was −0.32 (95% CI −0.57 to −0.06, I² = 87%, P = 0.01) and disability SMD was −0.44 (95% CI −0.71 to −0.17, I² = 89%, P = 0.001), both from low-quality evidence. When two outlier studies (Monticone 2013, Khan 2014) were removed, the pain effect weakened and lost significance (SMD −0.11, P = 0.07), while the disability effect, though smaller, remained significant (SMD −0.16, P = 0.001). | High |
| No benefit of CBT was maintained at follow-up | There were no statistically significant differences between CBT and comparison therapies for pain or disability at the 3-, 6-, or 12-month follow-up time points, suggesting effects may not persist beyond the intervention period. | High |
| CBT may improve fear avoidance and self-efficacy | Across 5 studies (505 participants), CBT was associated with reduced fear avoidance (SMD −1.24, 95% CI −2.25 to −0.23, I² = 96%, P = 0.002). Across 5 studies (1,060 participants), CBT improved self-efficacy (SMD 0.27, 95% CI 0.15 to 0.40, I² = 74%, P = 0.0008). Both estimates came from low-quality evidence with high heterogeneity. | Medium |
| CBT as an add-on to other therapy outperformed that therapy alone | In subgroup analysis, CBT plus a control intervention was better than the control alone for pain (SMD −0.67, P = 0.01) and disability (SMD −0.81, P = 0.003); after removing two outliers the pain effect lost significance (SMD −0.12, P = 0.09) while disability remained significant (SMD −0.13, P = 0.03). CBT versus active therapy alone showed no significant difference for pain or disability. | Medium |
| Educational background may influence who benefits | Subgroup analysis of sociodemographic characteristics suggested participants with a higher educational background (college grade or higher) may benefit more from CBT for pain relief (P = 0.04). | Low |
Immediately after intervention, pain SMD was −0.32 (95% CI −0.57 to −0.06, I² = 87%, P = 0.01) and disability SMD was −0.44 (95% CI −0.71 to −0.17, I² = 89%, P = 0.001), both from low-quality evidence. When two outlier studies (Monticone 2013, Khan 2014) were removed, the pain effect weakened and lost significance (SMD −0.11, P = 0.07), while the disability effect, though smaller, remained significant (SMD −0.16, P = 0.001).
There were no statistically significant differences between CBT and comparison therapies for pain or disability at the 3-, 6-, or 12-month follow-up time points, suggesting effects may not persist beyond the intervention period.
Across 5 studies (505 participants), CBT was associated with reduced fear avoidance (SMD −1.24, 95% CI −2.25 to −0.23, I² = 96%, P = 0.002). Across 5 studies (1,060 participants), CBT improved self-efficacy (SMD 0.27, 95% CI 0.15 to 0.40, I² = 74%, P = 0.0008). Both estimates came from low-quality evidence with high heterogeneity.
In subgroup analysis, CBT plus a control intervention was better than the control alone for pain (SMD −0.67, P = 0.01) and disability (SMD −0.81, P = 0.003); after removing two outliers the pain effect lost significance (SMD −0.12, P = 0.09) while disability remained significant (SMD −0.13, P = 0.03). CBT versus active therapy alone showed no significant difference for pain or disability.
Subgroup analysis of sociodemographic characteristics suggested participants with a higher educational background (college grade or higher) may benefit more from CBT for pain relief (P = 0.04).
Strengths
- First meta-analysis to evaluate the effect of CBT on chronic low back pain without restricting CBT intervention type or provider, and the first to investigate CBT's effect on self-efficacy
- Large pooled sample (3,003 patients from 20 RCTs) searched across five major databases (PubMed, EMBASE, Web of Science, Cochrane Library, PsycINFO)
- Followed PRISMA reporting guidelines, was prospectively registered with PROSPERO, used GRADE to rate evidence confidence, and applied a random-effects model with sensitivity and subgroup analyses
- Assessed multiple clinically relevant outcomes (pain, disability, fear avoidance, self-efficacy) at posttreatment and at 3-, 6-, and 12-month follow-up
Limitations
- Substantial variation between studies in CBT delivery format, duration, and provider qualifications made direct comparison difficult
- High statistical heterogeneity (I² often 74–96%) and predominantly low-quality evidence, with results sensitive to two outlier studies
- Inability to blind participants to treatment introduced potential performance bias favoring CBT
- Only English-language studies were included, raising the risk of language bias
- Follow-up periods were inadequate to firmly establish long-term benefit, and inconsistent reporting of clinical significance prevented firm conclusions about real-world clinical meaning
Key Takeaways for Patients
What This Means for You
- 01Cognitive-behavioral therapy (CBT) is a structured psychological approach that teaches understanding of pain, active coping strategies, and problem-solving for difficult situations.
- 02For chronic low back pain, CBT may offer small improvements in pain, day-to-day function, fear of movement, and confidence in managing pain right after a course of treatment.
- 03The benefits seen in this review were generally small and based on lower-quality evidence; for pain in particular, the advantage largely disappeared once two unusual studies were removed.
- 04CBT appears most helpful when added on top of another treatment (such as physical therapy) rather than used instead of it.
- 05There was no clear evidence that CBT's benefits lasted beyond the treatment period at 3, 6, or 12 months, so maintaining gains over time remains uncertain.
Read the Full Paper
Access the complete peer-reviewed study from Pain Research and Management
View Full Study