a "placebo" generally refers to an inactive control arm of a study of a drug. For studying something like meditation, you would use a different kind of control arm. The most common example is "treatment as usual". So half of ppl get meditation therapy for a given condition (say, depression), half get "treatment as usual" (pills and Or therapy). This has lots of problems as "treatment as usual" varies a lot from patient to patient, doctor to doctor

A better control is basically taking every aspect of mindfulness based therapy except for the mindfulness (i.e. You educate the patient about the condition, have a group therapy session with a healthcare provider, have at-home therapy assignments, etc). You standardize this "active control" and compare how patients do with this intervention compared to the mindfulness based intervention. This isolates the effect of mindfulness vs other potential confounding effects, and also should reduce physician to physician variability

In the one large study of mindfulness interventions using this active control, there was no measurable benefit to the mindfulness arm