At Cancer Decisions RWM has written two posts called “Kombucha Reconsidered”. After drinking kombucha for a while, he decided to stop. One reason was lack of evidence of benefit:
When I began investigated the actual medicinal properties of Kombucha tea, I thought I would be overwhelmed with information. Not so. For something that has been around for so long, there are only 38 scientific articles in PubMed on the topic of kombucha. Most of these are technical studies on the nature of the bacteria and yeast in the brew. Only a few of these are clinical.
In particular, no evidence of benefit for cancer:
But I am unaware of any credible data linking kombucha consumption to the prevention of either recurrences or metastases. (PubMed yields just two articles on the topic of kombucha and cancer, both of them negative.) This is a poor basis on which to make health decisions.
He also found two case reports, one from 1995, the other from 2009, where kombucha might have caused illness. In the 1995 the evidence is weak; in the 2009 report the connection is more plausible — but the sick person had HIV. The authors nevertheless generalize to everyone: “Consumption of this tea should be discouraged.”
This is a reason self-experimentation is important: So you can ignore inane statements in research articles. After I found that flaxseed oil improved my balance, I could ignore research that supposedly showed poor conversion of short-change omega-3 (in flaxseed oil) to long-chain omega-3 (used by the brain). Had RWM managed to measure the effect of kombucha on himself, he would have a vastly better basis for deciding whether or not it helped him.
This is also a reason that theory is important. John Tukey, the statistician, spoke of “gathering strength” when analyzing data. It is rare that a single body of data tells you how to analyze it, he said. (For example, what transformation to use.) You should use similar data sets to help decide. Scientific theory has the same effect. Before I started drinking kombucha, I didn’t have obvious digestive problems (unlike a friend) and my immune system seemed to work well. So it wasn’t easy to measure its effect. Yet I drink it and am untroubled by the evidence that worries RVM because I have a theory: the umami hypothesis (that we need a steady intake of bacteria to be healthy). This allows me to assess the effect of kombucha — whether it is likely to be good or bad — with the help of evidence from other bacteria-rich food (yogurt, natto, etc.) and much different data (the effect of bee stings on arthritis, hormesis, epidemiology, the effects of turmeric, etc.). Because the umami hypothesis appears to be true, apparently bacteria intake is beneficial — and kombucha has lots of bacteria.
Thanks to Tom George.