Mind-body exercise in mild cognitive impairment: benefit for cognition, with dose patterns that need confirmation

Design

Systematic review and meta-analysis of 27 randomised trials with restricted cubic spline dose-response

Population

2,034 adults aged 60 or older with mild cognitive impairment

Primary outcome

Global cognition and processing speed after mind-body exercise

Effect

Global cognition Hedges' g 0.67 (95% CI 0.49 to 0.78); processing speed 0.45 (0.3 to 0.6)

This meta-analysis pooled 27 randomised trials (2,034 people aged 60 or older with mild cognitive impairment) of mind-body exercise such as tai chi or qigong, and asked how dose related to cognitive effect.

There was moderate-certainty evidence of improved global cognition (Hedges' g 0.67, 95% CI 0.49 to 0.78) and processing speed (0.45, 95% CI 0.3 to 0.6). The weekly dose-response for global cognition was U-shaped. At 240 minutes a week, the relative low point, the pooled effect remained significant (0.37, 95% CI 0.19 to 0.55), and achieving benefit at that dose needed at least 11 weeks. Memory followed a similar pattern at very low certainty.

The authors say the dose findings need prospective validation, and weekly doses in the trials were heavily clustered, so a curve fitted across them is uncertain. The robust message is that regular mind-body exercise over a few months appears worthwhile.

• Consider recommending tai chi or similar mind-body exercise to older adults with mild cognitive impairment.
• Aim for a programme lasting at least about 11 weeks rather than a short course.
• Do not prescribe an exact weekly dose from this analysis; the curve is exploratory.
• Combine with other measures such as vascular risk control, hearing correction and social engagement.
• Record baseline cognition to track change.

Why it matters

There are few effective options for mild cognitive impairment, and this is low-risk and low-cost.

Don't overread it

Moderate certainty at best, with doses clustered and the dose-response curve exploratory; it does not show that exercise prevents dementia.

The statistics, in plain English

A Hedges' g of 0.67 is a moderate effect, and exercise trials cannot easily blind participants, which can inflate results. A U-shaped curve fitted to clustered doses can be an artefact, so the apparent low point at 240 minutes should not be taken as a recommendation.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.