NEW YORK – The composition of microbes living on people's skin or in their mouth can be used to estimate their age, give or take a few years, a new study has found.
The human microbiome changes rapidly during the first few years of life, and while it then settles down, the gut microbiome in particular is known to keep changing with age. This led researchers from the US and China to explore whether other human microbiome sites change with age, too.
They combined data from numerous studies totalling nearly 9,000 microbiome samples, including gut, mouth, and skin samples, from people living in the US, UK, China, and elsewhere. Using a machine learning approach, they found that individuals' skin microbiomes were the best predictors of age, as they could gauge people's ages within about four years, as they reported yesterday in the journal mSystems.
"This new ability to correlate microbes with age will help us advance future studies of the roles microbes play in the aging process and age-related diseases, and allow us to better test potential therapeutic interventions that target microbiomes," said co-senior author Zhenjiang Zech Xu, who was a postdoc at the University of California San Diego School of Medicine when the study was conducted and is now at Nanchang University in China, in a statement.
For their analysis, the researchers combined microbiome samples from 10 studies, which included 4,434 fecal samples, 2,550 saliva samples, and 1,975 skin samples. The samples came from people between the ages of 18 and 90 who had no history of inflammatory bowel disease or diabetes and who had not taken antibiotics in the preceding month.
The team used a random forest approach to regress the relative abundances of amplicon sequence variants from the various microbiome samples against the individuals' ages.
The gut microbiome was, as previously established, associated with chronological age. But the associations between the oral and skin microbiomes and age were even stronger, the researchers found. The skin microbiome could predict an individual's age to within an average 3.8 years, while the oral microbiome could predict it within an average 4.5 years and the gut microbiome to within an average 11.5 years.
The oral and gut microbiome of young people, those between 18 and 30 years of age, were more diverse and abundant than those of older adults over the age of 60. While the researchers noted some differences in the gut microbiome composition by sex, there was no such differentiation in the oral and skin microbiomes.
The changes in skin microbiome with age could be due to known aging-related skin changes, such as decreased sebum production and increased dryness, the researchers noted.
Next, they plan to use this machine learning approach to examine correlations between the microbiome and clinical conditions, including inflammation in autoimmune disorders. In the future, they said, this could form the basis of a noninvasive test of the microbiome to diagnose or determine risk for disease.
"The accuracy of our results demonstrates the potential for applying machine learning and artificial intelligence techniques to better understand human microbiomes," author Ho-Cheol Kim from IBM Research-Almaden said in a statement. "Applying this technology to future microbiome studies could help unlock deeper insights into the correlation between how microbiomes influence our overall health and a wide range of diseases and disorders from neurological to cardiovascular and immune health."