Nothing says relax better than a peaceful evening in front of a steamy cup of chamomile. Since thousands of years, humanity uses it as a natural remedy for a large amount of ailments, most notably hypertension, sleeplessness and to ease a flu-dominated night, like in my case recently.
Moved by curiosity, I took some time investigating what is scientifically known about the therapeutic effects of chamomile and their mechanism of action. The results, I must say, are interesting and conflicting. Let’s examine what I was able to gather from around the internet and in a couple of scientific papers.
Chamomile is a class of plants whose main representatives, at least for infuse-making, are Matricaria recutita (German chamomile) and Chamaemelum nobile or Anthemis nobilis (Common Chamomile). It can be found wild or cultivated. Its flower is actually a composite daisy-like sprout. The flowers are the tiny yellow corollae forming the central bulb.
The pleasant fragrance the chamomile flowers produce arises from a large set of compounds (more than 120) including in particular sesquiterpenes such as chamazulene and alpha-bisabolol, flavonoids and flavanoids, like Chrysin, and many others. Most of these compounds are not present in free-form, but are bound to sugars through fragile bonds that can be broken easily, for example by heating.
Some research on the effects of these compounds has been performed. In particular, Chamazulene has been found to have antioxidant properties, together with matricin, alpha-bisabolol, and apigenin. Chrysin appears to show anxiolytic effect in laboratory rats (see also here and here) but nothing has been said for humans yet.
Other experiments show that chamomile can have small antibacterial effect on the gut’s bacterial population, both in human and rat. This finding, however, is a mere hypothesis to explain changes in the excreted substances.
So, it appears that chamomile does indeed have relevant activity, and for what concerns anxiolytic effects, some evidence exists. Despite this, The National Institute of Health page for chamomile reports insufficient evidence for most of the claimed therapeutical advantages of chamomile: the report is “C: Unclear scientific evidence for this use”, with only one case (“post-operation sore throat”) where a conclusion has been reached as “D: Fair scientific evidence against this use”. This is a very important example on how evaluation of pharmacological effectiveness is performed. Even when evidence supports presence of therapeutic effect from a given compound or preparation, only a set rigorous tests performed on human subjects allows to finally grant recognition of therapeutic effectiveness (or lack of it). In the case of chamomile, tests have been mainly performed in rats and mice. Even a single successful (or unsuccessful) human test is not enough to grant A (strong positive evidence) or F (strong negative evidence) grades in the Natural Standard grading scale. The grade C refers specifically to
- Evidence of benefit from >1 small randomized trials without adequate size, power, statistical significance, or quality of design by objective criteria, OR
- conflicting evidence from multiple randomized trials without a clear majority of the properly conducted trials showing evidence of benefit or ineffectiveness, OR
- evidence of benefit from >1 cohort/case-control/non-randomized trials AND without supporting evidence in basic science, animal studies, or theory, OR
- evidence of efficacy only from basic science, animal studies, or theory.
Let’s examine the cases one by one.
The first case occurs when tests result in positive evidence (it works) but the test is not “statistically significant” meaning that it has been performed on too few subjects: for something to be considered working, it must present an effect that occurs with some consistency. If you test a substance Foo on a single sick person, and he recovers, it does not mean that Foo is a cure. That person could have recovered just because he was lucky or strong enough to recover, regardless of Foo. A better test would be: treat 200 sick people with Foo, and take also 200 sick people with no Foo treatment, then compare the recovery rate in the two groups. If in the first group 180 people recover, while in the other only 20 recover, there’s definitely a good point in favour for Foo being effective in curing that sickness. Statistical analysis allows you to decide which numbers of people can be considered strong evidence or not enough evidence for Foo being an effective cure.
The second case is when two or more tests produce conflicting results. For example if laboratory A sees a recover in its people using Foo, but laboratory B sees no recover. There could be many reasons for this. Improper testing could be one, and even if all tests are performed properly there could be additional factors we don’t know. Example: suppose that people at laboratory A had an unknown strong ease of recovery from that sickness (because they are immune for some biological reason), so the group without Foo medication recovers as well as the group without Foo. The conclusion for the laboratory A is that Foo has no effect, while laboratory B says that Foo has an effect. This is conflicting evidence, and must be resolved by checking more people, until a clear majority allows a unique conclusion to be drawn.
The third case is when there is evidence for recover, but there’s no evidence from known science, animal studies or theory able to explain the observed phenomenon. This can lead to a scientific breakthrough if a new biological mechanism is found and explained, but until then, it is not possible to say anything about the pharmacological validity. This point also raises the difference from cohort case, control case, randomized or non-randomized trial. It would be an interesting discussion, but it goes a bit outside of my current knowledge, and I am determined to learn more about the details in the future.
The fourth and last case is when evidence exists only because we infer it should work from what we known today of the human body’s mechanisms, but no actual test has been performed, or tests have been performed only on animals.
In the case of chamomile, as of today we cannot confirm officially and with strong evidence on humans that a pharmacological effect does exists, because all tests have been performed on animals, with the very few human trials available still insufficient to draw significant conclusions. This does not mean that the effect does not exists. It could exists, or it could not, and whatever the truth is we cannot put an approval stamp on it yet, because we haven’t tested enough. In agreement to the scientific method, unless something is demonstrated via evidence to hold, it is assumed not to hold. It’s like presumption of innocence in criminal trials: someone is assumed innocent until proven guilty from evidence. The other way around would be disastrous.
Now, we can probably claim that the pure fact of preparing chamomile and enjoying its pleasant fragrance has a relaxing effect, but that would be a psychological effect triggering internal biochemical actions inside our body, finally leading to a relaxed mood. Mozart could have the same effect. The point is, from the pharmaceutical point of view, the correct answer (as of today) to the question “does chamomile really relax ?” is “some evidence exists that it does, but it’s still not enough to say for sure.”