Article Text

Download PDFPDF

Significant methodological flaws limit conclusions drawn by authors of a recent PTSD mindfulness study
  1. Daniel J Lee1,2,
  2. Charles W Hoge3
  1. 1US Food and Drug Association, Silver Spring, Maryland, USA
  2. 2Uniformed Services, University of the Health Sciences, Bethesda, Maryland, USA
  3. 3Center for Psychiatry & Neuroscience, Walter Reed Army Institute of Research, Silver Spring, Maryland, USA
  1. Correspondence to Dr Daniel J Lee, US Food and Drug Association, Silver Spring, MD 20850, USA; Daniel.James.Lee{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

ABSTRACT FROM: OpenUrlCrossRefPubMed

What is already known on this topic

Of six major international post-traumatic stress disorder (PTSD) treatment guidelines, only two mention mindfulness-based treatments, and none recommend their routine use.1 ,2 Few mindfulness-based intervention studies exist for PTSD, and all have significant limitations in design, outcome measures and/or data handling.1 However, these interventions remain widely used. A core-component of PTSD treatment addresses autonomic hyperarousal, and many clinicians apply mindfulness techniques for this or in facilitating treatment for patients too avoidant for trauma-focused psychotherapies (TFPs).

Methods of the study

This was a 17-week RCT involving 116 US Veteran's Association participants, comparing group mindfulness-based psychotherapy against group present-centred psychotherapy. The primary outcome, change in symptom severity, was assessed using the PTSD checklist (PCL; range, 17–85 with higher scores indicating more severe symptoms). As a secondary efficacy outcome, the Clinician-Administered PTSD Scale (CAPS) was used. A unique strength of this study was the use of active-control, rather than wait list. Randomisation and psychotherapy fidelity measurement were also carried out well. Unfortunately, design problems included: (1) duration of individual sessions differed between intervention and control (2.5 vs 1.5 hours); (2) participants randomised to mindfulness received a full-day silent retreat that controls did not; (3) intervention and control groups appeared to be similar in overall use of psychoactive medication and prior psychotherapy, but medication classes and types of psychotherapy (trauma vs non-trauma focused) were not specified; (4) treatment groups included three participants with subthreshold PTSD, although this small number is unlikely to have significantly affected results.

What this paper adds?

  • This article is important because of its active-control condition (ie, a control condition which at least partially accounts for face-to-face interaction with a psychotherapist and other non-specific factors). The majority of psychotherapy controls involve the absence or delay of intervention as a control (ie, waitlist condition).

  • No difference between the treatment and control group existed: PCL 54.4 (51.2–57.6) vs 56 (52.9–59 and CAPS 49.8 (44.3–55.3) vs 50.6 (45.4–55.8), respectively. Differences of <2 points on total PCL and total CAPS are not statistically or clinically significant.3 ,4 These results were somewhat masked by the emphasis on the less clinically significant variable of change over time which was called clinically significant improvement. These outcomes 66.7 (52.1–81.2) (n=30) for mindfulness and 54.5 (41.4–67.7) (n=30) for control were created by combining pre/post-treatment outcomes, time and treatment×time interaction as fixed effects and the intercept and slope as random effects with an unstructured covariance matrix then comparing this against averages from other PTSD research studies. The merit and validity of this method is uncertain. However, the fact remains that the CAPS scores started and ended at almost identical points for both groups.

  • This study suggests that present-centred psychotherapy may have actually been superior to mindfulness-based psychotherapy, as it delivered identical total PCL and CAPS scores in roughly half the exposure time afforded to mindfulness (13.5 vs 26.5 hours).


  • Data reporting and analysis were questionable as the authors relegated a gold-standard interview-based outcome measure to a secondary outcome measure and choose to focus instead on clinically significant improvement using a less reliable patient-rated assessment. Interview-based assessments such as CAPS are widely seen as being more accurate and better reflective of total symptom burden.5 Instead of focusing on total symptom burden on outcome measures, the authors analyse statistical differences in slope of change which carries a high likelihood of not being clinically meaningful.

  • Mindfulness patients were encouraged to practice between sessions while controls were given no out-of-group assignments.

  • The primary outcome was a self-report PTSD measure that was not blind. As a whole, this study would be considered to be at moderate-to-high risk of bias using the Cochrane tool (

What next in research?

Non-inferiority studies of evidence-based trauma-focused psychotherapies compared with mindfulness are needed to determine if efficacy is comparable, as well as studies of whether mindfulness increases efficacy of trauma-focused treatments when used adjunctively. Active-control comparisons are optimal and should ensure equivalence in therapeutic time to reduce potential bias. Systematic bias assessment needs to be routinely considered in study design and evaluation of published research, using instruments such as the Cochrane bias tool.5

Do these results change your practices and why?

These results are not likely to change our clinical practice. This study had significant limitations in terms of design, data collection, data analysis and data reporting which undermine its conclusions. Mindfulness may provide one modality that supports the relaxation component of PTSD treatment but should remain no more than adjunctive to established evidence-based treatments.



  • Competing interests None declared.

  • Provenance and peer review Commissioned; internally peer reviewed.

Linked Articles