Systematic review of transcendental meditation

Dear Editor,

In the June 2004 issue you give a short report (p. 154) on a review¹ of the effects of transcendental meditation (TM) on cognition, which concluded the benefits were due to expectations. Ironically for a journal that reviews holistic systems of medicine, the main problem with the review is its inclusion criteria, which results in a fragmented understanding both of TM overall and its effects on cognition. This is seen both in studies that are included and excluded from the review.

One study² on the elderly, mean age 81, which was included in the review, found TM subjects differentially improved on three measures of cognition over 3 months, some significant and some trends, which are described as ‘largely negative’ in the review and report. As the review was only about cognition it could legitimately not report on findings that the TM group improved significantly on systolic blood pressure (P < 0.005) after 3 months and nurses’ mental health improvement ratings after 18 months (P < 0.025), which could conceivably be related to cognitive functioning. After 3 years 100% of the TM group were still alive compared with 65% in the relaxation group (P < 0.005). After 15 years cardiovascular disease was 44% less for the TM group (P < 0.03).³ Such long-term effects cannot be due to expectations and it is highly likely they are related to improvements in cognition as part of an overall improvement in health.

A replication study⁴ showed that over a 3-year period 29 subjects increased their IQ from 116 to 125 (P < 0.001) and 50 subjects increased their field independence (P < 0.005). Subjects were 23 years old at pretest and these measures do not normally show such marked improvements after adolescence. The study was excluded because it was not randomised and the subjects were practising the more advanced TM-Sidhi programme in addition to TM. Lack of randomisation may mean the results only apply to those particular individuals, which is unlikely, but a balanced review of TM and cognition needs to consider such findings.

There are other criticisms one could make but space does not allow.

References

1. Canter PH, Ernst E. The cumulative effects of Transcendental Meditation on cognitive function – a systematic review of randomised controlled trials. Wien Klin Wochenschr 2003; 115: 758–66.
2. Alexander CN, Langer EJ, Davies RI et al. Transcendental Meditation, mindfulness and longevity: an experimental study with the elderly. J Pers Soc Psychol 1989; 57: 950–64. [Abstract]
3. Walton KG, Schneider RH, Nidich SI et al. Psychosocial stress and cardiovascular disease 2: Effectiveness of the Transcendental Meditation programme in treatment and prevention. Behav Med 2002; 28: 106–23.
4. Dillbeck MC, Assimakis PD, Raimondi D et al. Longitudinal effects of the Transcendental Meditation and TM-Sidhi program on cognitive ability and cognitive style. Percept Mot Skills 1986; 62: 731–8.

Reply to letter from Spivack,

To include only randomised and controlled trials of an intervention is standard practice for a rigorous systematic review. Regarding the trial conducted by Alexander et al., our systematic review depends on the data reported in the original peer-reviewed trial report¹, not the rose-tinted version published by the Maharishi University of Management (MUM).² The original trial report clearly states that there were no significant differences between groups in post-test scores for the Dementia Screening Test. The TM group did perform significantly better than no treatment on the associate learning subtest but pairwise comparisons between TM and the two active control procedures were not statistically significant. Similarly, none of the pairwise comparisons between groups for the word fluency subtest, the colour word interference test and the object uses test reached statistical significance. The only test of cognitive function in which there was a statistically significant difference favouring TM over active controls was the overlearned verbal task, a non-validated test of dubious value administered at post-test only. The test relies on the fact that elderly subjects would have overlearned the original version of the US Pledge of Allegiance before the additional words ‘under God’ were added in 1954. Subjects were scored correct if they could recall the new wording. Our description of this study as ‘largely negative’ is therefore entirely justified.

Our conclusion that the positive effects observed in four of the 10 trials included in the review are probably due to expectation or similar non-specific effects applies only to outcomes of cognitive function. Such effects may or may not also apply to health variables such as blood pressure. The study by Dillbeck³ was excluded because it failed to randomise subjects between treatments and used a different intervention (TM-Sidhi): it also had no control group. The reported changes cannot therefore be separated from a learning effect or other non-specific factors.

References

1. Alexander CN, Newman RI, Langer EJ et al. Transcendental meditation, mindfulness, and longevity: an experimental study with the elderly. J Personal Soc Psychol 1989; 57: 950–64. [Abstract]
2. Chalmers R, Clements G, Schenkluhn H, Weinless M. Scientific Research on Maharishi’s Transcendental Meditation and TM-Sidhi programme. Netherlands: MVU Press, 1990. Collected Papers Vol. 4. Vlodrop
3. Dillbeck MC, Assimakis PD, Raimondi D et al. Longitudinal effects of the Transcendental Meditation and TM-Sidhi program on cognitive ability and cognitive style. Percept Mot Skills 1986; 62: 731–8.