Pilates, yoga and qigong each led a different outcome for cancer patients across 73 trials

Ask an oncology nurse what helps with the anxiety and you will get a recommendation, delivered with confidence, that has almost no comparative evidence behind it. Yoga, usually. Sometimes tai chi. The choice has been a matter of what the hospital happens to run.

Feng and colleagues, writing in Applied Psychology: Health and Well-Being, put five of these practices into one network meta-analysis, the method that ranks treatments never tested head to head by chaining them through a shared comparison: yoga, tai chi, qigong, pilates and dance, across 73 randomized controlled trials involving 5130 participants. All of them helped. None of them helped with everything.

Pilates ranked highest for quality of life, yoga for depression, and qigong for anxiety.

Three outcomes, three different winners

The split is the paper’s real contribution, because it turns a vague recommendation into a specific one.

Mind-body exercise significantly improved quality of life, reduced depression, and alleviated anxiety, with the anxiety effect largest at nearly 0.9 standardized units and quality of life smallest at around 0.56. Those pooled figures cover all five practices together.

Underneath them, the network meta-analysis revealed clear efficacy hierarchies, and the hierarchy changes depending which outcome you care about. That is a more useful answer than a single winner would be. Someone whose main problem is a racing mind at three in the morning has a different question from someone who cannot get through a day.

Reading the anxiety number carefully

The biggest effect in the paper is also the shakiest, and the two facts are connected.

Anxiety showed the largest movement, by 0.89 standardized units, in a range running from 0.37 to 1.40. That span covers everything from a moderate effect to a very large one, and a range that wide usually means a handful of small trials disagreeing with each other rather than many good ones agreeing.

Quality of life is the opposite case: 0.56, in a much tighter range. Smaller, better measured, more likely to survive the next trial. If you had to bet on one of these results replicating, it would not be the headline.

The curve that turns over

Dose-response analysis identified an inverted U-shaped relationship with an overall optimal dose of 490 MET-min/week for quality of life, and intervention-specific optimal doses ranging from 300 to 520 MET-min/week.

A MET-minute, short for metabolic equivalent minute, multiplies how hard an activity is by how long you do it, with one MET being your body at rest. Yoga and qigong sit around 2.5 to 3 METs, so an hour of either earns roughly 150 to 180. About 490 a week is therefore three sessions of an hour or so, which is a realistic prescription rather than an aspirational one.

An inverted U means the benefit rose to that point and fell after it. In a cancer population that shape is easy to believe: fatigue is the defining symptom of treatment, and a program that runs past what someone can recover from stops being restorative. This is one of the few areas of exercise medicine where too much is a live concern rather than a theoretical one.

What a ranking probability can and cannot establish

Pilates ranked highest for quality of life (SUCRA = 84.80%), and those SUCRA values, the surface under the cumulative ranking curve, are ranking probabilities. They answer the question of how likely each option is to be the best, given the network.

Two things follow. Options separated by a couple of percentage points are not meaningfully different, and pilates at 84.80 against qigong at 83.27 is well inside that margin for the outcomes they do not lead on. And an option tested in few small trials can outrank one tested in many, because a large uncertain estimate produces a high probability of being best.

Read the hierarchy as a reasonable starting point for a choice, not as a demonstration that one practice beats another.

What a patient can actually do with this

The practical version is unusually concrete for this literature: pick the practice that matches the symptom bothering you most, aim for something like three hour-long sessions a week, and stop short of the point where it costs you more energy than it returns.

For anyone in or after treatment, the route runs through the oncology team rather than around it. Cancer chemotherapy uses medicines to destroy cancer cells, and what is safe during that depends on blood counts, surgical recovery and where the disease sits, none of which a meta-analysis knows about a particular person.

What has changed is the nature of the conversation. The question is no longer whether gentle exercise is permitted during cancer treatment. It is which kind, and how much, and the evidence now has something to say about both.