Skip to main content.

Journalists, don’t just follow your gut when covering the microbiome

Journalists, don’t just follow your gut when covering the microbiome

Picture of Daniel  McDonald
[Image: NIH/NHGRI via Flickr.]
(Photo: NIH/NHGRI/Flickr)

Every time you take a number two, you're flushing a wealth of information about the microbes that form your own personal ecosystem, also known as your gut microbiome. The amount of data, per gram of stool, is equivalent to a thousand 256GB iPhones. These microbial inhabitants of yours, which produce a suite of molecules that your body thrives on, play a critical role in your health. That’s why these organisms have become an intense area of research. And unlike your human genome, it’s possible to change your microbiome.

Using the data from DNA sequencing, we can compare and contrast samples to ask whether some microbial groups are, for instance, associated or correlated with a particular disease. Some studies go much further, and attempt to see whether the microbiome may be causing a disease. For instance, you can take fecal matter from an obese human, introduce it to the mouse, and the mouse will get fatter. There are a few other diseases where we have observed a causative association with the microbiome in mice, including Parkinson’s, multiple sclerosis, kwashiorkor and others.

Remember, correlation is not causation

The microbiome field is full of correlative or associative stories, and it's easy for journalists to fall into the trap of mistaking correlation for causation when reporting on exciting research findings.

For instance, within the , run through the University of California, San Diego’s School of Medicine, we observe an association between the number of types of plants people self-report eating in a week and the composition of their microbiome. This association is exciting as it provides evidence that a diverse diet is important to the organization of microbes in the human large intestine. But, we cannot say that a diverse diet is healthy, nor can we say definitively that a diverse diet is why the microbiomes we observed were different.

As scientists, journalists, readers and consumers, it’s imperative to remain skeptical about new findings and to be careful when reading between the lines of evidence. There are many exciting stories in the microbiome field, including those providing evidence that extreme diet changes will alter the microbiome, or that the microbiome is implicated in Autism Spectrum Disorder. But while these studies lend evidence, they don’t establish the mechanisms that may be involved.

As a rule of thumb, studies that demonstrate causation are uncommon in the microbiome field, particularly in humans. As a journalist, if your impression is that a result is causative, or implying a cure, it may be useful to a microbiome researcher unaffiliated with the study to seek further clarification. For causative studies in mice, it’s not safe to extrapolate the results to humans or other host types, as mice and humans differ, particularly in their immune systems.

So what can you do as a journalist?

What can you do right now to report appropriately to your readers about their microbiome? Here are a few useful contextual questions that may be worth considering at the start:

  • What are the claims being made, and are the claims supported by peer-reviewed research?
  • Who is making the claims, and do they have a conflict of interest?

Beyond this, there are a few types of questions that could avoid pitfalls when reporting on microbiome science. First, it may be useful to consider the study design. For instance, was it a study that tested for an association between two groups of individuals? Or was it a longitudinal study that looked at how the microbiome changes over time in response to some variable? If the study design is longitudinal, was the observed effect consistent across individuals? These are just two types of common study designs, in addition to the gold standard: double-blind randomized controlled trials.

Second, many microbiome studies suffer from small sample sizes (the number of people in the study), which can lead to false discoveries. There does not exist a perfect sample size, as that is dependent on the variable being tested. More subtle effects, like sex, require large sample sizes to detect a difference in the microbiome, whereas strong effects, such as antibiotics, require smaller sample sizes. Unfortunately, effect sizes are not well understood in the microbiome, which makes determining appropriate sample sizes hard. Generally speaking, caution may be warranted when a study uses tens of individuals per group, although deeply studied individuals (i.e., “n of 1” projects) can be quite informative if done with care.

Finally, talking directly with the researchers — and researchers not involved in the study — can be helpful for how to emphasize the importance of a study. If you’re having trouble finding an impartial independent expert to speak with, groups like the at UC San Diego (the American Gut Project, for which I work, is affiliated with the center) may be able to suggest researchers to . In fact, many major universities are creating “microbiome cores” and these groups should be able to connect you to researchers in the field for comment. The is also an excellent source of up-to-date information about the microbiome, including fecal microbiome transplants and probiotics.

Remember, it’s always smart to be skeptical about language that implies causation or cures. When in doubt, be sure to follow up with a microbiome scientist unaffiliated with the study under discussion for a second opinion.

Leave A Comment


Join us on April 23 for insights and suggestions from physician and author Dr. Sunita Puri that will deepen your understanding and coverage of palliative care and end-of-life issues. Sign-up here!


левитра купить

Dapoxetine 60 mg