Can you trust gut microbiome test results? Maybe not

Finding out which gut microbes a person carries may not be as simple as many companies claim.

Seven direct-to-consumer microbiome testing companies each got three identical fecal samples but gave different results on the intestinal microbes present, researchers report on February 26 in Communication biology.

The results highlight differences between companies which claim to give consumers insight into their gut health. This is important because consumers may take probiotics they don’t need, change their diet in harmful ways, or even get fecal transplants based on inaccurate microbiological test results, say researchers at the University of Maryland in Baltimore and the National Institute of Standards and Technology in Gaithersburg, Maryland. (Molecular geneticist Scott Jackson has since left NIST and is now a consultant for companies whose products involve microbes.)

The NIST team created a fecal standard by putting several people’s stools into a blender to create a homogenized sample. “We know that the biology is identical in all of these samples,” NIST microbiologist Stephanie Servetas said Feb. 25 during a press briefing. “What this material is supposed to do is really tell how reproducible the results are, whether between companies or within a company, but it won’t be able to tell us who was closest to the right answer.”

Some companies matched the microbes they identified in all three samples. But one company produced a drastically different result on one of three samples tested. He classified the two similar results as “healthy” and the outlier as “unhealthy.” Others identified many of the same types of bacteria in replicate samples but showed varying amounts of those bacteria. The differences between the companies’ results on the standard sample were similar to the variability between samples from different people.

NIST began selling the fecal standard to companies last year to use for calibration and quality control, which could lead to improved testing methods in the future, Servetas said. The goal is not to force companies to adopt the same methods or stifle innovation, she said, but “there should be minimal guidelines and controls” that would make results more consistent.

One sample, three results

One company had large inconsistencies in the genera of microbes (indicated by different colors, gray being unspecified) that it identified in three replicates of a standard sample. The company classified the first two results as healthy and the third as unhealthy.