| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
LETTERS TO THE EDITOR |
Maharishi University of Management, Fairfield, Iowa
Davidson and colleagues’ recent paper in Psychosomatic Medicine (1) used the extensive work on the functional neuroanatomical substrates of emotion to illuminate possible mechanisms of meditation practice. Their work, originally based on EEG studies (2, 3) and supported by current PET and fMRI research (4), has associated positive and negative affect with frontal brain activation (5, 6) . Specifically, increased activation in left dorsolateral and ventral medial prefrontal cortices, as well as in the amygdale, has been observed during certain forms of positive emotion and more generally during approach-oriented disposition/behavior toward rewarding stimuli. Right prefrontal cortex activation has been observed during certain forms of negative emotion and during withdrawal-oriented disposition/behavior away from aversive stimuli.
Their recent paper reported three findings resulting from mindfulness meditation practice: 1) "significant increases in left-sided anterior activation, a pattern previously associated with positive affect;" 2) "significant increases in antibody titers" (a measure of immune functioning); and 3) that the "magnitude of increase in left-sided activation predicted the magnitude of antibody titer rise."
A close examination of their paper suggests that these conclusions are inaccurate. First, significantly greater left-sided activation was only observed at central sites (C3/C4) not frontal sites (pg. 566). C3 and C4 are over motor cortices (7), which play little, if any, part in emotion processing. Thus, greater "left-sided anterior activation" was not observed in this study. Also, the authors never discussed why left-sided motor cortex would be significantly activated as a result of mindfulness meditation. They also do not discuss why frontal activation, which has been associated with approach/withdrawal dispositions and positive/negative affect, did not change as a result of mindfulness meditation.
Second, Figures 3 and 4 are confusing. The emotion research strongly suggests that positive emotions lead to increased left-sided activation; this is seen in Figure 3. However, increased left-sided activation is also seen in Figure 4, which is in response to negative emotion induction. This obvious contradiction was not discussed.
Third, Figure 6 in this paper shows the relation between changes in so-called asymmetric "anterior" activation (which was really C3/C4 asymmetry) and antibody titer. Figure 6 contains a significant outlier (first point on the lower left-side of the figure). Correlation and regression statistics are very sensitive to outliers, which vastly inflate correlation values (8). Visually, if the outlier was removed, the high correlation appears as though it would drop from significance.
This research was well-designed and investigates an important research question—long-term effects of meditation practice in daily life. However, the conclusions from this study seemed to reveal: 1) increased left-sided activation in motor cortex with practice of mindfulness meditation, suggesting increased motor activity in the right side; 2) immune response may increase with mindfulness meditation, but the change in immune function does not appear related to changes in brain functioning.
There is increasing interest by the public, the medical community, and funding agencies in complementary and alternative medicine modalities in general and particularly in meditation. Careful analysis and interpretation of data are required to ensure maximum progress in our understanding and appropriate utilization of meditation for improving individual and collective health. We offer these observations in the spirit of cooperation to build a reliable comprehensive knowledge base to guide mind/body research.
REFERENCES
University of Wisconsin, Madison, Wisconsin
We thank Travis and Arenander for their comment because it provides us with the opportunity to clarify a number of issues that were clearly misunderstood in their reading of our article. They raise a number of questions about the findings we reported in the Davidson et al. (1) study on the effects of mindfulness meditation on brain and immune function.
First, the authors claim that our findings of greater left-sided activation were found only at the central sites (C3/C4). This is simply false. As Figure 3 of the article illustrates, the same basic effect was also obtained at the anterior temporal sites (T3/T4). The authors then assert that the C3/C4 electrodes are over the motor cortices, "which play little if any part in emotion processing." First, the electrodes reflect activity more broadly, including the premotor frontal region and somatosensory cortex, which Damasio and colleagues have strongly implicated in emotion processing. More to the point, in other research cited in our article, we have published findings indicating that positive emotion is associated with increases in left-sided activation at the C3/C4 sites. Reference 11 in our article presents these data (2). Travis and Arenander further claimed that we "did not discuss why frontal activation, which has been associated with approach/withdrawal dispositions and positive/negative affect, did not change as a result of mindfulness meditation." Again, this claim is untrue. On page 569 of the original article, the entire second paragraph of the Discussion section is devoted to exactly this issue! We noted in that paragraph that the fact that we did not find significant increases in dispositional positive affect in the meditation group may be related to our failure to detect significant increases in left prefrontal activation. We further suggested that if the duration and/or intensity of the intervention were increased, we might have then been able to detect increases in both positive affect and left prefrontal activation.
Travis and Arenander then suggest that Figure 4 is misleading because in it we show that for the meditators, the negative affect induction is associated with left-sided activation. Figure 4 is not a comparison of the positive vs. negative affect induction conditions. Rather, it depicts the changes over the course of treatment in activation asymmetry in response to the negative affect induction. Most importantly, the third paragraph of the Discussion section focuses explicitly on this issue. We note that other work of ours and other investigators has found that increased left-sided activation in response to negative affect inductions is associated with better regulation of negative affect (3). Finally, the authors raise a question about a possible outlier in the scatter plot presented as Figure 6. This is precisely why the scatter plot was included rather than just reporting the r value and p value of the correlation. One of the limitations of our study that we discuss in the Discussion section is the small sample size that adversely affects the robustness of our findings in several different ways. It is indeed the case that elimination of a single data point in the scatter plot presented in Figure 6 could alter the correlation in different directions depending on which data point was eliminated. We could also significantly strengthen the correlation by eliminating a single data point (most extreme point in the lower right quadrant). We devote a paragraph in the Discussion section to the various limitations of this initial study, and the very first limitation we discuss is the small sample size of this study. Clearly, these findings require replication and extension in a larger sample.
We appreciate the opportunity to clarify these points and share with Travis and Arenander their conclusion that "Careful analysis and interpretation of data are required to ensure maximum progress in our understanding and appropriate utilization of meditation for improving individual and collective health."
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
