No Evidence for CBT and Other “Conversion Disorder” Therapies

By David Tuller, DrPH

I have recently written critically about the CODES trial of cognitive behavior therapy as a treatment for the phenomena that have long been called “psychogenic non-epileptic seizures” but that some neurologists and psychologists are trying to rebrand as “dissociative seizures.” (Many people understandably find it offensive to be told they have a “psychogenic” disorder; perhaps the term “dissociative” comes across as less dismissive.) Now Cochrane has published a review that relates directly to this issue–“Psychosocial interventions for conversion and dissociative disorders in adults.”

The new review is an update of a 2005 version. Its stated objective was “to assess the beneficial and harmful effects” of the psychosocial interventions in question, several of which were forms of CBT. The review included 17 studies published up until last year, with a total of 894 participants. The 2010 pilot study for CODES was included, although CODES itself was published too late. The Cochrane review found that the studies were generally of poor quality and produced little or no evidence of benefits from a range of psychosocial interventions.

Here are the conclusions of the Cochrane review:

The results of the meta‐analysis and reporting of single studies suggest that there is lack of evidence regarding the effects of any psychosocial intervention of conversion and dissociative disorders in adults. It is not possible to draw any conclusions about potential benefits or harms from the included studies.

The study’s main authors were mental health experts at Danish psychiatric and psychological centers. It is therefore not suprising that the language used in the review suggests they take the purportedly psychogenic nature of these conditions at face value. Here’s the abstract’s “background” section:

Conversion and dissociative disorders are conditions where people experience unusual neurological symptoms or changes in awareness or identity. However, symptoms and clinical signs cannot be explained by a neurological disease or other medical condition. Instead, a psychological stressor or trauma is often present. The symptoms are real and can cause significant distress or problems with functioning in everyday life for the people experiencing them.”

Noting that the reported symptoms “cannot be explained by a neurological disease or other medical condition” implies that such an explanation does not exist. The other possible explanations—that the underlying pathophysiological causes have been missed or that the science has not yet identified them—do not appear to have been considered. But that’s a shortcoming of the studies included in the review as well, given that they were mostly based on hypothesized and unprovable mechanisms like the “conversion” of psychological distress into somatic complaints.

It does seem like some psychiatrists and neurologists have sought to supplant the conversion and psychogenesis framings, as the use of the term “dissociative” suggests. In CODES, the investigators do not present the medical events they are studying as psychogenic or examples of conversion disorder. But if understanding of the phenomenon of unexplained seizures has evolved beyond the reductive constructs of conversion and psychogenesis, it is unclear why a form of CBT based on those constructs would still be appropriate. The CODES paper does not explain why the investigators decided to test an intervention grounded in a strictly psychogenic model that they seem to believe is outdated.

Had the null results in CODES for the primary outcome of seizure frequency at 12 months been included in the Cochrane analysis, they would certainly have reinforced the review’s conclusions. In CODES, participants who did not receive CBT designed to reduce seizures experienced greater seizure reduction at 12 months than those who did receive the specialized intervention, although the results were not statistically significant.

For the CODES investigators, the results from the pilot were convincing enough to proceed with a huge trial. Yet in assessing the 2010 pilot study, the Cochrane review described the findings like this: “Cognitive behavioural therapy may have little or no effect on reducing physical signs [monthly seizure frequency] at end of treatment.” Cochrane rated the certainty of the evidence produced by the pilot study as “low” per the GRADE system, which defines “low” this way: “further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.”

We have since had that further research—CODES–and can say with a certain amount of confidence that CBT does not reduce seizure frequency at 12 months. (I have previously addressed why the secondary outcomes are basically meaningless, even though they were the basis on which King’s College London hailed CODES as a great success.)





7 responses to “No Evidence for CBT and Other “Conversion Disorder” Therapies”

  1. Richard Vallee Avatar
    Richard Vallee

    Do you find it weird that this update was completed right along with the conclusion of the largest experiment on the topic being published? It seems logical to wait for those results and include them, being the definitive trial and all.

    I find that very weird. In a deliberate way. Sure is one way to bury null results. And considering some of the people involved have a pattern of doing just that, with intent, it seems difficult to think of any other explanation considering it is all but guaranteed that had the CODES findings been a resounding success, it would absolutely have been included in this review.

    Frankly this is beyond a pattern, it’s clearly a deliberate strategy. But all the i’s are dotted and all the t’s are crossed so I guess that’s all good.

  2. Sally Jones Avatar
    Sally Jones

    Good article, as usual. Thank you for your continued advocacy.
    I have a cartoon about you on my website at and a couple of new cartoons in the ongoing collection. Best wishes to you.

  3. Patricia Price Avatar
    Patricia Price


    “Positional Spinal Cord Compression Linked With FM

    In 2002, Heffez also advocated viewing the cervical cord more dynamically with MRI. His suggestion was presented in the context of sorting out another complex presentation, fibromyalgia (FM), at a small meeting of the National Fibromyalgia Research Association, in Portland, Oregon. His analysis found that symptoms of FM and cervical myelopathy overlapped for many patients in FM. Further, he noted that when he surgically decompressed severe, comorbid myelopathy, FM symptoms (allodynia, fatigue, central sensitization, pain, dyscognition, sleep disturbance) often abated.14 Similar observations were made for FM combined with Chiari malformation.15

    Figure 2. Sagittal neutral, flexion, and extension MR cervical spine images in a patient

    To say that Heffez’s message was received with mixed responses would be overly generous. A furor built quickly to relegate it to near oblivion. Nary a clinician wanted to consider the specter of surgical intervention for 6 to 10 million American patients with FM.

    In light of the above findings, beginning in 2003, radiologists working with the Pacific Rheumatology Associates began obtaining flexion-extension cervical spine MRI routinely for patients without additional charge. Positional cervical cord compression (PC3) was defined as clear, visually-confirmed abutment of the cervical cord, with a canal diameter measured at 10 mm or less
    (Figure 2). (Of note, the typical cervical spinal canal can average 14 to 16 mm in men and 13 to 15 mm in women, depending on disc level and patient size.) Compression or abutment of the cord was often intermittent, and consistent with prior studies, with PC3 much more evident in extension rather than in flexion.

    From a database that now exceeds 3,000 patients, a random two-month experience was reported in The Journal of Pain in 2008.16 Of 107 referrals to a suburban rheumatology office, 53 had FM by 2001 American College of Rheumatology (ACR) classification criteria,17 32 had an autoimmune or connective tissue disease, and 22 had chronic widespread pain (CWP) without sufficient allodynia (tender points) to confirm FM. PC3 was identified by flexion-extension C-spine MRI in 71% of the FM group and in 85% of the CWP group. Also, in only 15 of 52 patients identified with PC3 (21%) could the cord compression be visualized on the traditional, neutral sagittal MRI view. Thus, 80% of patients in this study with PC3 would not have been diagnosed using the MRI protocol available to most practicing clinicians.”

    “To say that Heffez’s message was received with mixed responses would be overly generous. A furor built quickly to relegate it to near oblivion. Nary a clinician wanted to consider the specter of surgical intervention for 6 to 10 million American patients with FM.”

  4. Patricia Price Avatar
    Patricia Price

    Aaaannnddd a Swedish study **from this year** confirming the overall effect of the findings discussed above.

  5. lucie webb Avatar
    lucie webb

    There is a gap between medical and dental treatments emanating from the base of the brain. A gentle approach to treatment could be all that is needed.
    See the web pages of the late Darden Burns, especially her comments on Oral System Balance
    and also the web page of Pauline Bresnik-Snasdell

  6. Joan Crawford Avatar
    Joan Crawford

    Thanks for sharing this David. I’ll take a wee look.

    There is evidence of patients labelled with FND, dissociative seizures etc with no appreciable trauma history. So, somatising theory falls flat here with this subgroup. Patients grouped together in trials are likely highly heterogeneous without good quality structured assessment and stratification so it’s no surprise null results are commonplace. Frustrating to see resources squandered and time wasted.

    CODES bizarrely didn’t treat or test patients often disabling trauma reactions (it was rather generic unfocused CBT not related to the trauma focused CBT used often successfully for PTSD reactions). So, clinicans like myself are left none the wiser whether treating patients trauma (if present) has any impact on seizure frequency and quality of life. It’s odd that this specifically has not been assessed via clinical trials – considering the transmission of psychological distress into physical symptoms is the cornerstone of their theories. My hunch, having worked clinically with many patients, is that reducing a trauma reaction will help patients to cope better with the FND/seizures. In a similar way to patients with chronic pain plus PTSD. Treat the PTSD and the patient can cope a lot better managing their pain/seizures. Their pain/seizures are still there but hopefully will have improved QoL. Conflating the two only annoys and alienates patients from medical and psychological support. Sad. And if there is a subgroup of patients who improve greatly with therapy they need to be well defined so that they can be identified and have access to good quality care to recover.

    CODES also failed to comment on the very low physical functioning score identified in the study. Makes me think that there is a lot more to this.

    Joan Crawford
    Counselling Psychologist

  7. cracksoftwar Avatar

    you can kill virus by using this antivirus . its free for you.