Research on meditation produces results that are debated more than the results of drug trials on comparable questions. The disagreement is largely methodological, and the obstacles are specific.
Participants always know what they received
A drug trial can give an identical-looking tablet to both groups, so neither participants nor assessors know who received what.
Nobody can be unaware of whether they spent eight weeks in a meditation programme, which makes participant blinding impossible in principle rather than merely difficult.
Expectation therefore differs systematically between groups, and expectation influences the self-reported outcomes these studies most often use.
There is no agreed inactive comparison
A placebo tablet is inert. No equivalent exists for a practice involving time, attention, instruction and often a group.
Waiting-list controls receive none of those, so any difference includes the effect of doing something structured with other people.
Active control conditions have been developed to match those non-specific components, and studies using them typically report smaller effects than waiting-list comparisons do.
The intervention is not one thing
Practices grouped under the same heading differ substantially in instruction, in what attention is directed towards, and in what the participant is told to do with distraction.
Programmes also vary in length, in teacher training and in how much home practice occurs, and adherence to home practice is often self-reported.
Pooling such studies risks averaging across interventions that are not comparable, which complicates any summary of the evidence.
Outcomes are mostly self-reported
Stress, wellbeing and mindfulness itself are measured by questionnaire.
Some mindfulness questionnaires ask participants to rate qualities the training explicitly teaches them to notice, so scores can shift as vocabulary changes rather than as experience does.
Physiological outcomes avoid this but are further from what people want to know, and the two do not always move together.
What this means for reading the field
Poor blinding and weak controls tend to inflate measured effects, so the strongest claims typically come from the weakest designs.
Better-controlled trials generally find smaller but non-zero effects for some outcomes, which is a more modest and more durable position than either enthusiasm or dismissal.
The pattern is not unique to meditation. It applies to any behavioural intervention that cannot be disguised, which is a limitation of the method rather than of the practice.
Adverse experiences during intensive practice are also documented, which is one reason people with a history of trauma or psychosis are advised to seek qualified guidance rather than beginning intensive practice unsupported.