Movement Practices for Chronic Pain Compared: Yoga, Tai Chi, Pilates and Exercise

What is "chronic pain"?

A Cochrane overview defines chronic pain as pain lasting beyond normal tissue healing time, generally taken to be 12 weeks. [1] Most of the head-to-head research below is in chronic non-specific low back pain, meaning back pain with no clear structural cause. Other pain conditions have their own pages, such as yoga for fibromyalgia and chronic pain and yoga for arthritis.

Where does exercise in general stand?

• Cochrane overview of exercise for chronic pain (2017). It covered 21 reviews, 381 studies and 37,143 participants, and included yoga, Pilates and tai chi among many exercise types. [1] The authors judged the quality of the evidence low, with favorable effects that were mostly small to moderate and not consistent across reviews. [1] Most adverse events were soreness or muscle pain, which reportedly settled after a few weeks. [1]
• Cochrane review of exercise for chronic low back pain (2021). It included 249 trials and found moderate-certainty evidence that exercise treatment is more effective than no treatment, usual care or placebo. [2] Against other conservative treatments the effects did not meet the authors' threshold for a clinically important difference. [2] They also noted 79% of trials had risk of performance bias because exercise cannot be blinded. [2]
• American College of Physicians guideline (2017). For chronic low back pain it advises starting with nonpharmacologic treatment, including exercise. [3] Tai chi and yoga are on its list of options, graded as low-quality evidence. [3] Pilates is not named in the abstract's list.

Head-to-head: what each practice's own reviews found

Two network analyses of similar trials reached different rankings. That is a strong sign that rankings of "best" practice are not reliable. The more cautious analysis said directly that evidence does not establish a preferred modality or an optimal dose. [9]

Why no practice can be declared the winner

• Most trials cannot be blinded. People know whether they are doing yoga, tai chi or Pilates. In the yoga review, outcomes were self-assessed, and the authors thought it unlikely that blinded comparisons would find a clinically important benefit. [4]
• Comparators differ. A trial against a waiting list is easier to "win" than one against another exercise. Comparator type was linked to effect sizes in the 2026 network analysis. [9]
• Small, varied trials. The Cochrane overview noted that most included studies had fewer than 50 participants in total. [1] The tai chi analysis flagged substantial heterogeneity and small sample sizes in several trials. [6] See understanding evidence in complementary health.
• Preference and access. The yoga review authors said the choice of yoga instead of no exercise or another exercise may depend on availability, cost, and participant or provider preference. [4]

NCCIH summarizes it this way: for low back pain a large amount of yoga research suggests a slight benefit, while for other conditions the evidence looks promising but small. [10] NCCIH adds that these approaches have good safety records, which does not mean they are risk-free for everyone. [11]

How to choose

Pick the practice you are likely to keep doing, that your clinician agrees with, and that you can afford and reach. Questions to ask any teacher: What experience do you have with pain? Can I modify everything? Are you comfortable with me skipping poses? Read the individual guides: yoga for low back pain, Pilates for back pain, tai chi, qigong, and the broader comparisons yoga vs tai chi and yoga vs Pilates.