Monday, May 20, 2013

An In Depth Look at a Mindfulness and Stress Study

My former classmate at the University of Hartford, Tonya Jacobs, had an article described in the Huffington Post titled Mindfulness Meditation Could Lower Levels Of Cortisol, The Stress Hormone.  It described her study which originally appeared in the Journal of Health Psychology.  It talks about how a study was conducted of those who participated in a three month meditation retreat where it says mindfulness was measured along with cortisol levels in saliva as a measure of stress.  If found that cortisol (a hormone that strongly indicates stress) levels were decreased from the beginning of the retreat to the end as mindfulness increased.  They warned that there was no control group.

From past experience I know that the news media often leaves out important details of a study and can be bad at interpreting the results of the studies.  I thought I would take a look Jacobs et al. (2013) original article which has a lot more detail to allow one to replicate the study.  In the methods sections they state that there was a treatment group of 30 for those who participated in the retreat and a control group of 30 who were wait listed for the retreat using stratified random sampling to control for any potential confounding variables such as BMI (Body Mass Index which is weight divided by height squared and is a crude measure of obesity), handedness and IQ.  The wait listed group did receive the mindfulness intervention after the treatment group.  

The authors do not discuss if there is a difference between the wait list and non waitlist groups in cortisol or mindfulness levels especially during the time where the wait list group was not receiving the intervention.  This may not have been feasible during the study as the individuals were probably scattered all over the US and measurements of cortisol and mindfulness could not be taken without great cost.  Participants were paid during the retreat.  Three participants had incomplete data and were excluded from the data analysis.  

The results showed no overall effect of cortisol decrease from pre to post measurements but did significantly decrease for BMI.  Mindfulness significantly increased between the pre and post.  The pre and post cortisol levels were significantly negatively correlated with mindfulness as measured by a 37 item questionnaire which was previously validated.  Negatively correlated means that as mindfulness increases cortisol levels decrease.  This effect was still significant after adjusting for age and BMI.

The authors acknowledge that this study is correlational and does not establish a cause and effect relationship between meditation, mindfulness, and stress.  The article in the Huffington Post seems to suggest the same thing by stating that there is no control group.  Because of the difficulty in doing this type of research, the need for more of theses types of studies is established.  Experimental studies with a well defined control and treatment group with all other confounding variables are adjusted for are ideal for establishing cause and effect relationships.  The ideal of research is seldom met.  When the situation is less than ideal, converging validation with many different methods is needed to accomplish the same cause-effect relationship.  


**Related Posts**


ADHD, Genetics, and Causality: A Chicken-Egg Problem



Cause and Effect, Slip Slidin' Away 






Draft Logic


3 comments:

  1. You have several errors and misinterpretations in your blog.

    1) Your blog states, “Three participants had complete data and were excluded from the study” This is not accurate. No one was excluded if they had complete data. As stated in the study, only participants with complete data across both cortisol and mindfulness assessments were included (N = 57).

    2)Your blog states, “Participants were paid during the retreat” This is not completely accurate. As described in the study, participants paid for their room and board, but were compensated for the testing sessions.

    3)This blog may have missed the main point of the study:

    The main finding was the novel association between cortisol and the measure of mindfulness, regardless of whether participants had gone through the meditation retreat or not (i.e., it occurred both before and after the retreat).

    The validity of this relation does not require a control group.

    The fact that cortisol and mindfulness changed together over time strengthens the credibility of the association.

    5) The data were not intended to imply a causal relation, but take the first important step in showing an association. The next step is to use experimental methods to discern whether there are unidirectional or bidirectional relations between the two.

    ReplyDelete
    Replies
    1. This comment has been removed by the author.

      Delete
    2. Thanks for your comment. I was not attacking their findings. Jacobs et al., in the discussion section of their article, acknowledge that their findings are preliminary. It is often important to establish a correlational relationship before establishing a cause and effect relationship. I corrected the first complaint. As for point three I was criticizing the Huffington Post's interpretation of the Jacobs et al. study not the authors'.

      Delete