The Yoga RCT

[Editor’s note: With this post we welcome back Dr. Apoorva Chandra, who previously wrote about Ayurveda. He returns with a post about a questionable clinical trial of yoga. Welcome back!]

This isn’t even a significant study that needed any attention in the first place to be honest. But, one fine evening, the chief scientist of WHO decided to tweet about this study.

The Ministry of AYUSH, which legitimises pseudoscience and gives pseudoscience (including homeopathy) a legal platform in India, was also tagged in the tweet. To give the benefit of doubt and because of her excellent track record, I would like to believe that this tweet was an error or was done in haste and not intentional. However, Dr. Swaminathan seemed to defend the tweet/study in follow up tweets despite the flaws in the study that were rightly pointed out by many. “We should see the full paper and then draw conclusions. I agree that methodology is important and all RCTs are not of a similar standard. We should focus on robust evidence generation to back claims of efficacy of any product” said the chief scientist of WHO responding to one of the very valid criticisms of the study on Twitter. That sounded like a reasonable suggestion although I wish Dr. Swaminathan had seen the full paper and realised how badly designed it was before tweeting about it to reach conclusions that are not supported by the study’s claims. Anyway, let me take this suggestion forward and look into the full study and see if there’s anything on it that is new, ground-breaking, or noteworthy.

To begin with, let me clarify, I’m not against yoga. Yoga (sans all the woo, hype, mysticism, supernatural claims, and the accompanying pseudoscience) would probably be beneficial as a part of a healthy lifestyle as a form of exercise when done within reasonable limits. It is the ‘cure for all’ claims that I’m against. Regular physical exercise has shown to be beneficial for multiple conditions and for general health as a whole in multiple validated studies and there’s no doubt about it. Yoga isn’t anything special or an exception to that. If the claim is that a specific type of yoga is beneficial to a specific condition or disease process then the study has to be very specific and explicitly include these specifics – something that I have not seen so far. Vagueness seems to be the bread and butter of alternative medicine studies.

I am against misrepresentations like those found in this study, claiming yoga to be different from and superior to exercise just because you call it ‘yoga’.

I am also against being vague about the definition of yoga and trying to market it as some kind of magic cure for everything or a secret science now being revealed.

Let us have a closer look at the study in question.

Methodology

The study seems to measure following markers to claim ‘yoga’ is superior to exercise:

• BMI
• blood glucose levels
• HbA1c
• lipid levels
• IL6
• TNFα
• TBARS
• Adiponectin
• PTGIS
• sleep quality

As a clinician, measuring BMI, blood glucose levels, HbA1c, and lipid levels is understandable but the clinical relevance of the rest of the so-called inflammatory markers in diabetes is still not fully understood and not something routinely measured in clinical practice. Using their values as surrogate markers is therefore largely unhelpful. I will talk about ‘sleep quality’ later because this RCT is going to get more interesting soon.

This was a randomised control study involving 2 groups. Of course, due to the nature of the study, blinding of participants was not an option.

But look at the 2 groups. They only defy logic at every step from here on.

Group 1/No yoga “was advised to do simple physical exercises for 50 minutes for 5 days in a week”. There is no explanation what these ‘simple physical exercises’ were or how they were ‘advised’. However, the interventions with Group 2 are very interesting.

Group 2/Yoga group: “was advised yoga intervention and were instructed to practice it for a period of 50 minutes for 5 days in a week”. All the subjects in group 2 were trained by a yoga teacher. During their first session, each participant was educated on diabetes, risk factors, its complications, and the importance of yoga in maintaining glycaemic control and preventing the complications due to diabetes. In addition, they were given training on yogasanas using a separate yoga module for 30 minutes and static loosening exercise for 20 minutes. This training was given to the subjects individually and they practiced it in the supervision of the yoga teacher till they could do the module on their own. Later they were advised to practice the same activity for 5 days per week at home.

The yoga module (given as supplement 2) included quick relaxation techniques beginning with loosening exercises for the legs, spine, neck and shoulders and then asanas such as Thadasana, Trikonasana, Vajrasana, Konasana, Patchimothasana, Uttanapadasana, Sarvangasana, Matchyasana, Salabasana, and Pranayama including abdomen breathing, Nadisudhi, Kabalbhati, Sitali, and Brahmari relaxation technique. A copy of the module was given to them with clear instructions.

They were also instructed to contact the yoga teacher through telephone for any clarifications related to the asanas as they practiced at home. Telephone calls were also made by the researcher once in a fortnight to motivate the subjects for continuing yoga practice/exercise and also to increase compliance with the yoga practice and exercises among the study subjects. Very few required more than one session for learning the asanas and most were able perform them completely on their own. After the first session, few subjects had confirmed the correct procedure of the asanas with the yoga teacher by making video calls while they were practicing. The events such as hypoglycemia during the study period was also recorded.

A pause for thought

Let’s do a thought experiment: Bring in two teams, calling it an RCT between Cricket and Cricket Pro. Cricket team is advised to go and practice for 5 days a week for the next 3 months before the match. However, nobody cares once they are ‘advised’ and they are left on their own or nobody knows how they are doing for 3 months.

Cricket Pro team gets a coach, each member gets a personal trainer, are encouraged to carry on with different techniques, have regular compliance checks, etc. Conduct a match at the end of three months and declare Cricket Pro is better than Team Cricket as they won. Isn’t this more than obvious to everyone? What exactly are they trying to ‘prove’?

I am really not sure how they can be serious and call this a reasonable methodology for an RCT. I fail to understand why and how this methodology, with obvious differences between the two groups, was designed for an RCT with any goal other than to misrepresent something. Or is this the general lack of understanding of how RCTs are done? In both cases I’m worried.

And what about the sleep quality?

A quick word about ‘sleep quality’ in the yoga group. Sleep quality was measured via survey using a qualitative scale. There are various biases and confounders in play in this methodology and this is an extremely subjective and unreliable way of measuring something like sleep quality especially in a study like this.

So what can we conclude?

So, does this study prove what it claims to be proving? Absolutely NOT. This is a study with glaringly-obvious flawed methodology designed to present yoga as superior to exercise with inadequate data on the non-yoga group.

The study goes into details of the statistics of these different markers and contains different tables and graphs, but I’m stopping here because it is pointless to apply statistics and graphs when the methodology is almost hilariously bad.

Other points to consider:

• The study was funded by the Department of Science and Technology, Government of India, Science and Technology in Yoga and Meditation (SATYAM) project
• A WHO Collaborating Centre for Research Education and Training in Diabetes was also a part of the study
• This is in agreement with many other studies which show the benefits of regular exercise.

Is it a completely useless study? What else could it suggest? Possibly not. Although it is flawed, it seems to be suggesting regular exercise and stretching could be beneficial in a variety of health conditions, including diabetes, without demonstrating yoga brings anything extra to the table. Unfortunately the biggest concern is the WHO chief scientist, who holds a position of authority, tweeting about this study – which has in turn has been shared multiple times, especially by people who have keen interest in promoting pseudoscience on social media, as a ‘proof’ or ‘evidence’ for their other baseless claims because it has been shared by none other than the chief scientist of WHO.

It is really unfortunate that flawed studies like these are used to market ‘yoga’. It is important to appreciate yoga for what it is and call out misinformation and misleading claims in the name of yoga. While yoga may have real health benefits, the associated pseudoscience and ‘cure for all’ claims are actually detrimental to health. The real benefits of exercise or ‘yoga’ has been hijacked by numerous gurus and mystics promoting pseudoscientific and supernatural claims and unfortunately yoga has become a buzzword to mislead people into accepting claims that are not backed by science. ‘Yoga’ also has a brand value which drives many financial interests in disregard for the actual science. The present Indian government is enthusiastic about promoting yoga on an international level, which may not be completely harmful in itself but the government clearly lacks insight into the pseudoscience it intentionally and unintentionally promotes with this.

Lastly, we hope doctors, scientists and people who hold responsible positions review what they share on social media and not add more fuel to the already problematic pseudoscience industry.