Are at-home microbiome tests worthwhile?

Commercial microbiome home tests often promise insight into health, disease, or even the early detection of conditions. That sounds appealing, but the available research shows that the science is far from ready for this. The underlying biology is interesting and is being actively studied, but a reliable, validated home test does not yet exist.

One study examined whether a model based on seven gut-bacteria genera can identify rheumatoid arthritis (RA). The best model achieved an AUC of approximately 71-75%, a measure of diagnostic accuracy where 100% is perfect and 50% amounts to guessing. That is better than chance, but a clinically useful test generally requires an AUC of 90% or higher. This model is therefore not ready for use in a home test, let alone for making a diagnosis.

A rarely discussed technical problem makes the situation even more complicated: the way in which microbiome data are processed before analysis has a large influence on the outcome. Whether a laboratory log-transforms raw counts or works with relative abundances can substantially change the calculated accuracy. Results from two different labs or tests are therefore difficult to compare, and a home-test result depends heavily on technical choices that the consumer neither knows about nor can evaluate.

For cervical cancer, the vaginal microbiome is being investigated as a possible additional biomarker. This is purely exploratory research: no validated self-test based on microbiome data yet exists. The study in question mentions combining microbiome information with protein biomarkers as a promising research direction, but concrete accuracy figures are lacking.

In Alzheimer's research, gut flora analysis forms part of the PREDICTOM study, a comprehensive clinical screening programme. Results are not yet available, and the analysis takes place in a healthcare setting, not at home. In addition, a study protocol has been published for home-based modulation of the gut microbiome in COVID-19, but again no results are yet known. None of these applications currently yields a scientifically substantiated home test. For a healthy microbiome, you do more with enough fibre and fermented food than with a test.