Activity of the medial prefrontal cortex after psycho-spiritual healing (Baldwin et al., 2016).
Everything we do and feel and experience changes the brain
. Psychotherapy, juggling, taxi driving, poverty, reading, drugs, art, music, anger, love. If it didn't we'd be dead
. Why should prayer be any different? The trick is to accurately determine the structural or physiological changes that are unique to a specific activity. And when assessing the effectiveness of clinical interventions, how the changes compare to an adequately matched control intervention. Plenty of high profile studies have failed to do that, including a recent one on emotionally focused therapy
I feel bad about criticizing a study on the neural correlates of healing prayers
. I'm not one of those smug atheists
who lord their intellectual superiority over the unwashed religious masses. Certain atheist organizations claim
they're all about promoting scientific literacy and a secular worldview. But I think these New Atheists
are detrimental to science literacy, since they alienate the vast majority of the population.
So why am I blogging about a prayer intervention for depression? It's not to sneer at the authors. And it's especially
not to sneer at the participants, who were recruited from Houston-area churches. My interest is the unholy alliance between brain imaging and a psychological intervention with no control condition. As I've said before
...neuroimaging studies of psychotherapy that have absolutely no control conditions are of limited usefulness. We don't know what sort of changes would have happened over an equivalent amount of time with no intervention. More importantly, we don't know whether the specific therapy under consideration is better than another form of psychotherapy, or better than going bowling once a week.
Healing Prayer, Trauma, and Forgiveness
This is especially true for a treatment that is based on faith and a strong belief that the intervention will work —
a Christian form of prayer focused on forgiveness and psycho-spiritual healing (PSFH). A prayer minister “led the subject through three different phases: (1) a prayer of forgiveness for the perpetrator of the hurtful event; (2) a prayer of blessing on the perpetrator; and (3) a prayer to heal the emotional damage caused by the traumatic event.”
Study design for the 6 week healing prayer intervention (Baldwin et al., 2016).
The 18 participants had moderate to severe levels of depression on the Hamilton Depression Scale (HAM-D
). Oddly, post-traumatic stress disorder (PTSD
) was not assessed before or after the intervention. This was a major weakness, given that the purpose of the intervention was to forgive the perpetrator of childhood abuse and to heal from emotional trauma. In this sense, PSFH is akin to more formalized psychotherapies such as forgiveness therapy
It's no surprise that a non-randomized, unblinded prayer intervention in religious persons resulted in dramatically reduced HAM-D scores in the 14 participants who completed the study (11 of whom were available for a one year followup).
Who am I to criticize a practice that helps suffering people? I won't do that.
What I will do is point out difficulties in task design that make it nearly impossible to interpret some aspects of their fMRI study. The task used a symptom provocation
paradigm using 3 key words to evoke memories of the traumatic event (15 seconds) and feelings of the traumatic event (15 seconds), separated by a 2 second blank screen.2
Is it possible to separate traumatic memories from the feelings they evoke, and to switch between them on such short notice? Certain therapies (such as prolonged exposure
) are designed to do just that. The authors stated that anecdotally, this appeared to be the case here as well:
In this and our previous study, subjects frequently mentioned informally that PSFH results in a separation of the traumatic memory and associated feelings: while the memory remains intact, it no longer associates with traumatic feelings.
Activity of the precuneus to Bad Feelings was higher before psycho-spiritual healing (Baldwin et al., 2016).
It is, however, difficult to interpret a 23 voxel decrease in precuneus
activity in 14 subjects as a reflection of such a complex therapeutic change, especially since this brain region is involved in both self-referential processing and episodic memory retrieval
But to be even more fair, the authors listed ten caveats
to their admittedly preliminary study.3
When all is said and done, how can this study reveal ANYTHING about the neural correlates of healing prayer?
Or in this case, nothing fails like a non-randomized, unblinded, not-placebo-controlled fMRI study of prayer. Or of any other intervention, for that matter.
Johnson SM, Moser MB, Beckes L, Smith A, Dalgleish T, Halchuk R, Hasselmo K, Greenman PS, Merali Z, & Coan JA (2013). Soothing the threatened brain: leveraging contact comfort with emotionally focused therapy
. PloS one
, 8 (11).
Also see two blog posts
by Dr. James Coyne.
These Bad Memory/Feeling blocks were also compared to Neutral Memory/Feeling blocks that evoked memories and feelings about a neutral topic (e.g., the weather). This is the pre/post contrast shown in the first figure of the post.
To shorten and paraphrase the overly honest Limitations section of Baldwin et al. (2016)
Baldwin, P., Velasquez, K., Koenig, H., Salas, R., & Boelens, P. (2016). Neural correlates of healing prayers, depression and traumatic memories: A preliminary study Complementary Therapies in Medicine, 27, 123-129 DOI: 10.1016/j.ctim.2016.07.002
- the number of subjects was small (n=14)
- recruitment was largely done at churches, which might affect generalizability
- individual minister effects could not be ruled out
- there was no control population receiving an alternative therapy
- only subjects who completed the study were included, which may have skewed the results
- life events such as changing employment status, marriage stability, family, health, and economic changes were not assessed
- possible confounding effects between the role of PSFH and intercessory prayer for the participants by others [NOTE: some of us may discount this as a confounder]
- cannot rule out an effect of being exposed to the task in the MRI twice
- demand characteristics — participants answered worse at the beginning and better at the end to fulfill researcher’s expectations
- outcomes were rated by non-blinded observers