Mindfulness-based cognitive therapy is widely used to help people prevent recurrent depression, but a new individual participant data meta-analysis is examining a question that has received far less attention: whether the intervention might worsen depressive symptoms for some patients. Published in JAMA Network Open, the study brings together participant-level information from randomized clinical trials to investigate deterioration rather than focusing solely on average improvement. Its central concern is whether mindfulness-based cognitive therapy is associated with an increased likelihood of depression becoming more severe during or after treatment.
The question matters because clinical trials commonly report mean changes in symptom scores. An average reduction in depression can suggest that a therapy is beneficial overall, yet averages may conceal a smaller group of participants whose symptoms remain unchanged or become worse. By analyzing individual participant data, researchers can examine the distribution of outcomes more closely, compare the number of people who deteriorate across treatment groups, and investigate whether particular baseline characteristics are linked to worsening symptoms. This approach can provide a more clinically relevant assessment of both benefit and risk.
Mindfulness-based cognitive therapy, often abbreviated MBCT, combines mindfulness practices with techniques derived from cognitive behavioral therapy. Participants are typically taught to observe thoughts, emotions and bodily sensations without immediately reacting to them, while also learning to recognize patterns of negative thinking that can contribute to depressive relapse. The treatment was developed in part for people with a history of recurrent depression and is commonly delivered through structured group sessions, guided exercises and home practice. Rather than attempting to eliminate distressing thoughts, MBCT aims to change how individuals relate to them.
The therapy’s potential relationship with worsening depression is scientifically complex. Mindfulness exercises can increase awareness of internal experiences, including sadness, anxiety, intrusive memories or self-critical thoughts. For many people, this awareness may support emotional regulation and reduce automatic patterns of rumination. For others, particularly those with severe symptoms or limited support, focusing attention inward could initially feel destabilizing or intensify distress. A careful analysis must therefore distinguish temporary discomfort from clinically meaningful deterioration and determine whether worsening occurs more frequently than would be expected in an appropriate comparison group.
Randomized clinical trials provide the strongest framework for evaluating that question because participants are assigned to treatment or control conditions rather than choosing between them. Randomization helps balance known and unknown factors that could influence outcomes, such as illness severity, previous treatment, age or coexisting psychiatric conditions. The new analysis pools data from these trials at the level of individual participants. Unlike a conventional meta-analysis, which generally combines summary statistics reported by each study, an individual participant data meta-analysis can apply consistent definitions and statistical models across trials and can explore differences among participants and treatment settings.
A key challenge is defining what counts as worsening depression. Researchers may use changes in validated symptom scales, thresholds for clinically significant deterioration, loss of remission or the emergence of severe symptoms. Different definitions can produce different estimates of risk. A technically rigorous analysis must account for baseline scores, the timing of assessments, missing data and differences in the instruments used by the original trials. It must also consider whether participants who discontinue treatment are more likely to have experienced difficulty, since excluding them could make an intervention appear safer than it is.
The analysis is particularly relevant to the way psychological treatments are communicated to patients. A therapy can have a favorable average effect while still carrying a possibility of harm for a subset of participants. Reporting only mean symptom reductions may leave clinicians without enough information to explain that uncertainty during informed consent. Conversely, identifying any worsening episode as evidence of treatment-related harm could overstate risk if symptoms fluctuate naturally or if deterioration occurs at similar rates in comparison groups. The distinction between association and causation is therefore essential: an increase in symptoms during MBCT does not automatically demonstrate that mindfulness caused the increase.
The work by Simon B. Goldberg of the University of Wisconsin-Madison and Willem Kuyken of the University of Oxford addresses this issue through a collaboration spanning clinical psychology, psychiatry and quantitative research. By combining randomized evidence, the investigators can examine whether deterioration is more common among people receiving MBCT than among participants assigned to control conditions. They may also be able to assess whether the pattern varies according to treatment format, participant history or the level of depressive symptoms at enrollment. Such subgroup analyses are informative but must be interpreted cautiously, because dividing participants into many categories can produce apparently important findings that are statistically unstable.
The study’s focus on adverse outcomes reflects a broader shift in mental-health research toward balanced evaluation. Psychological interventions are often described primarily in terms of effectiveness, while harms, nonresponse and treatment dropout receive less attention. A comprehensive assessment should measure all three: who improves, who does not improve and who becomes worse. This is especially important for depression, a condition in which symptoms can change over time and in which clinical deterioration may carry serious consequences. More precise evidence could help practitioners identify patients who need closer monitoring or additional support while participating in MBCT.
The meta-analysis does not imply that mindfulness-based cognitive therapy is broadly dangerous, nor does the research question alone establish that it causes worsening depression. Its purpose is to test the possibility with data from randomized trials and to clarify how often deterioration occurs relative to control conditions. The results will be most useful if they present both the potential benefits and the full range of outcomes, including uncertainty around risk estimates. For patients and clinicians, the practical message is that mindfulness-based treatment should be offered with appropriate screening, realistic expectations and ongoing attention to symptom changes rather than treated as universally suitable or universally risk-free.
Subject of Research: Individual participant data meta-analysis of mindfulness-based cognitive therapy and worsening depressive symptoms.
Web References: https://doi.org/10.1001/jamanetworkopen.2026.29946
References: Goldberg SB, Kuyken W, and colleagues. Study published in JAMA Network Open; full article details and author list were not provided in the supplied material.
Keywords: Mindfulness-based cognitive therapy, MBCT, depression, depressive symptoms, worsening symptoms, individual participant data, meta-analysis, randomized clinical trials, psychotherapy, mental health, treatment safety, clinical psychology.
