We often hear that supplements are not scientifically validated because they have not passed through ‘Double-Blind, placebo controlled’ studies.
This is used as an excuse to scoff at generations’ worth of experience.
What is a DOUBLE BLIND study?
In short, when medicines are tested, a group of people who
are all suffering from a particular problem are selected. These people are
divided into at least two groups: one group will be given the medication
and the other group will NOT receive the medication, but an inactive placebo
instead.
In the ideal situation, neither the people being tested, NOR
the clinicians administering the doses know who is receiving what. Because
neither the patient (single blind) nor the clinician (double blind) know who is
receiving what, this kind of trial is called ‘Double Blind’. ‘Placebo
controlled’ means that one group is actually being given a ‘medication’ but it
contains NO active ingredients (as opposed to receiving nothing, for example).
In this way, it is hoped to minimise unmeasurable effects
such as HOPE which might have such a positive effect on the patients that they
improve through the body’s own efforts rather than the medication. Such a
positive reaction is well-known as the ‘placebo effect’ and can influence
results to the point that they may no longer be regarded as ‘scientific’.
Scientists are aware of how easy it is to view results
differently, depending on what you believe the outcome should be. A simple but
well-known example of this is the glass half-full/half-empty scenario. The aim
of double blind, placebo controlled studies is to produce unbiased results.
Unbiased Results?
Recording daily results and measurements is not difficult.
Interpreting them, is!
It is absolutely amazing how differently each body reacts to medication and even food. For example, a recent study on Blood Glucose (1) found that
1) The reaction between participants to a standardised meal varied greatly (up to 70%), and was determined by aspects such as gut flora, genetics, amount of sleep, enzymes present, etc.
2) The reaction which each participant had when given the exact same, standardised meal more than once, varied by an average of 30%.
Other factors which may affect the actual data include:
- The clinician’s expectations, mood on the day, technique in obtaining results and many more
- The participants’ mood, expectations and willingness to follow the protocol exactly
- The participants’ interpretation of how they feel: is pain better or worse, is this an (un)usual headache, is blood-pressure up through effort or medication, are these ‘palpitations’ from excitement or a side-effect?
- The weather, the friendliness and ease of access of the location (or not) and so many more
Once the data have been collected, they need to be organised and interpreted. This is where the REAL problem lies.
Research of any kind does not come cheap, and there are many ways in which researchers can (and do) please their sponsors to ensure future funding:
- Design the research with a specific result in mind (rather than to find the true answer)
- Select data which is pleasing and discard any which does not ‘fit’ the intended result
- ‘Massage’ data to obtain results which support the sponsor’s aims and claims
- Use innovative techniques (=cheating) to create graphs which look impressive but convey an incorrect impression (only a handful of people with access to the actual data can tell).
- Use ‘statistics’ to create a false impression
- An example of misrepresenting a totally insignificant result: in a random population of 5000 people, 2 people may get condition X per year. Using THIS medication on a test group of 5000 people, only 1 person got condition X. This is an improvement of 100%. YES!! One HUNDRED PERCENT improvement. But wait, there’s more…..By using THIS medication, your RISK of contracting condition X is REDUCED by HALF!!
- And many, many more!
- An example of misrepresenting a totally insignificant result: in a random population of 5000 people, 2 people may get condition X per year. Using THIS medication on a test group of 5000 people, only 1 person got condition X. This is an improvement of 100%. YES!! One HUNDRED PERCENT improvement. But wait, there’s more…..By using THIS medication, your RISK of contracting condition X is REDUCED by HALF!!
- And many, many more!
There are multiple examples of such techniques being used in the gold-standard scientific methodology of Double-Blind studies.
What is maybe even worse, are the cases where negative effects are KNOWN by the manufacturer early on during the development and production stages and are purposely covered up. Such examples include- Vioxx (Merck), an NSAID (Non-Steroidal Anti-inflammatory) which ended up killing upwards of 38 000 (and possibly as many as half a MILLION) people before being withdrawn from the market(2,3).
- Avandia (GSK), a diabetes drug which lead to possibly as many as 100 000 (one HUNDRED THOUSAND) cases of stroke, heart attack or heart failure (4)
- Paxil (GSK), an anti-depressant that caused INCREASED rates of suicide in teenagers
- Bextra (NSAID, Pfizer), Risperdal (Anti-psychotic, Johnsson & Johnsson) and many others prove that these are not isolated cases. (5)
