Patient-friendly summary
If you read nothing else
Bottom line
Several exercise types (motor control, Pilates, resistance, Tai Chi/Qigong, yoga) all help reduce chronic neck pain short-term versus no exercise, with no clear winner among them and uncertain long-term benefit.
Moderate evidencePublished
Evidence hierarchy
Study participants
Adults aged 18–70 with chronic neck pain lasting 12 weeks or more
Full research — for clinicians and curious readers
Study Summary
This systematic review of systematic reviews (umbrella review) summarized the evidence on five exercise types — motor control exercise (MCE), Pilates, resistance training, traditional Chinese exercise (TCE: Tai Chi/Qigong), and yoga — for adults with chronic neck pain lasting at least 12 weeks. Across 25 included reviews (221 RCTs, 17,321 participants with overlap not accounted for), the authors found low to high certainty evidence that all five exercise types produce short-term improvements in pain compared with non-exercise controls, and that all except resistance training also improve disability short-term. When exercise types were compared head-to-head with one another, or assessed over the intermediate/long term, results were conflicting or showed no clear advantage of one type over another. The authors conclude no single optimal exercise type can be recommended, and note the included reviews varied widely in methodological quality.
Key Findings
| Finding | Detail | Impact |
|---|---|---|
| All five exercise types reduced pain short-term versus non-exercise controls | Narrative analyses showed positive short-term effects on pain for MCE, Pilates, resistance training, TCE, and yoga compared with non-exercise controls (usual care, education, waitlist), with low to high certainty of evidence. | High |
| All exercise types except resistance training improved short-term disability versus non-exercise controls | MCE, Pilates, TCE, and yoga showed positive short-term effects on disability compared to non-exercise controls; for resistance training, results on disability were varying with low certainty in the intermediate/long-term and no significant short-term meta-analytic effect (SMD = -0.91, 95%CI -2.22 to 0.39). | High |
| No single exercise type was clearly superior to another | When exercise types were compared head-to-head, results were conflicting or showed no advantage; MCE was the exception, showing positive short-term effects on pain (SMD = -0.25, 95%CI -0.38 to -0.13) and disability (SMD = -0.36, 95%CI -0.52 to -0.20) versus other exercises. The authors conclude no optimal exercise intervention can be recommended. | High |
| Intermediate/long-term effects were uncertain | In the intermediate/long-term, results were mainly no effect or varying for both pain and disability, whether compared to non-exercise controls or other exercises. For yoga, no long-term results were available at all. | High |
| Methodological quality of included reviews was often low | Of 25 reviews, 5 were high quality, 7 moderate, 8 low, and 5 critically low (13 of 25 rated critically low to low). The main AMSTAR-2 shortfall was a non-comprehensive literature search (only 13 of 25 scored 'yes' on item 4). The authors also reported finding numerous data errors when spot-checking SR/MA data against original RCTs. | Medium |
| MCE showed a notably large short-term pain effect versus non-exercise controls | Pooling two SRs using a non-exercising comparator gave a large short-term effect for MCE on pain intensity (SMD = -1.69, 95%CI -2.73 to -0.64; I2 = 5%) and disability (SMD = -2.26, 95%CI -3.13 to -1.39; I2 = 0%), though these estimates come from few reviews. | Medium |
Strengths
- First systematic review of systematic reviews on various exercise types for chronic neck pain, to the authors' knowledge
- Conducted by a large group of experienced researchers following PRISMA and Cochrane Back- and Neck Group recommendations
- Comprehensive multi-database search (Ovid MEDLINE, Embase, Cochrane Library, SportDiscus, Web of Science) with no restrictions on year, comparator, or language
- Formal quality (AMSTAR-2) and certainty-of-evidence (GRADE) appraisal applied throughout
- Prospectively registered protocol (PROSPERO CRD42022336014) with no protocol deviations
- Overlap between reviews quantified using corrected covered area (CCA)
Limitations
- More than half of included reviews (13 of 25) were of critically low to low methodological quality, downgrading certainty of evidence
- The authors relied on data as presented in the included reviews and reported encountering numerous data errors when spot-checking against original RCTs
- High overlap of original studies across reviews for some exercise types (e.g., TCE CCA 41%, MCE high overlap), and for TCE there were more reviews (seven) than underlying RCTs
- Substantial heterogeneity in exercise style, dose, duration, and comparator groups limited clinical recommendations; optimal dose could not be determined
- Chinese-language systematic reviews were excluded due to lack of reading competence, which may have affected results
- No long-term data were available for yoga; intermediate/long-term findings overall were sparse and inconsistent
- Mind-related outcomes (e.g., quality of life, mood) were not assessed, only pain and disability
Key Takeaways for Patients
What This Means for You
If you have long-lasting neck pain, doing exercise generally helps reduce pain in the short term more than doing no exercise at all.
No single type of exercise — motor control/deep-neck exercises, Pilates, resistance training, Tai Chi/Qigong, or yoga — clearly beats the others, so you can reasonably choose a type you enjoy and will stick with.
Benefits are clearest in the short term (up to about 3 months); the longer-term picture is less certain because fewer studies followed people that far.
The best choice of exercise is one made together with your clinician, based on your preferences, since sticking with the program appears to matter for results.
These findings come from reviews of varying quality, so they should be read as encouraging but not definitive.
Read the Full Paper
Access the complete peer-reviewed study from BMC Musculoskeletal Disorders