A person's gut bacteria and the colony of microbes that live in the body and on the skin may serve as a unique identifier, much like a fingerprint, researchers said.
The study led by Harvard University is the first to investigate just how identifiable people are based on their bacteria, which can vary substantially based on a person's age, diet, geographic location and overall health.
"Linking a human DNA sample to a database of human DNA 'fingerprints' is the basis for forensic genetics, which is now a decades-old field," said lead author Eric Franzosa, research fellow in the Department of Biostatistics at Harvard.
"We've shown that the same sort of linking is possible using DNA sequences from microbes inhabiting the human body -- no human DNA required."
Scientists found that stool samples were particularly reliable. Up to 86 percent of people could be identified by their gut bacteria after one year.
Skin samples were less reliable. About one third of those could be matched to a person after one year, said the study in the Proceedings of the National Academy of Sciences.
But even if the samples could not be matched, there were very few false positives. In most cases, either a match was made or it was not, but rarely did it identify the wrong person.
The study was based on a pool of 120 people, among the 242 who have donated their stool, saliva and skin samples to the Human Microbiome Project, which maintains a public database for researchers.
A computer science algorithm was used to establish individual codes based on the donors' microbiomes. These codes were then compared to the same people's samples collected during follow-up visits, as well as to a pool of strangers.
Researchers said their study shows it is possible to match human microbiome samples across databases.
But they also raised the issue of ethics, warning that the practice could expose sensitive personal information such as the presence of a sexually transmitted infection, which could be detected from the microbiome without the subject's own DNA or consent.
"Although the potential for any data privacy concerns from purely microbial DNA is very low, it's important for researchers to know that such issues are theoretically possible," said senior author Curtis Huttenhower, associate professor of computational biology and bioinformatics at Harvard.