You’ve heard that saying, “age is just a number.”
It turns out that there’s a lot of truth to it. And, as the saying suggests, that chronological number many of us try to avoid mentioning on our birthdays is really just a number — and far less relevant than we assume. The real age we should focus on is our biological age, which is influenced by factors as diverse as genes, lifestyle, behavior and environment. Scientists say that this age — not the number that gets you into bars at 21, or lets you apply for Social Security at 62 — more closely predicts our lifespan and healthspan and is, therefore, a more accurate reflection of our true age.
But, so far, scientists have found determining a person’s biological age to be extremely difficult, if not impossible.
Now, a team of scientists report that the accurate prediction of biological age is becoming more feasible, thanks to artificial intelligence and new, vast publicly available datasets. The researchers summarized the current work on deep aging clocks in a recent issue of Trends in Pharmacological Sciences.
According to Alex Zhavoronkov, Ph.D, founder and CEO of Insilico Medicine, people are good at using images, videos, voice and even smell to guess another person’s age. But, AI can guess ages even more accurately, he said in a news release.
“Deep neural networks can do it better and we can now interpret what factors are most important,” said Zhavoronokov. “Very often when someone looks older than their chronological age, they are sick. A trained doctor can guess the health status of a patient just by looking at him or her. At Insilico we developed a broad range of deep biomarkers of aging that can be used by the pharmaceutical and insurance companies, as well as by the longevity biotechnology community. In this paper we describe the recent progress in this emerging field and outline a range of non-obvious applications.”
The researchers suggest that there’s not just one indicator — or biomarker — that can tip off a person’s biological age, but a combination of these different predictors. AI is suited for this task because it is adept at sorting through all the available data and determining the right combination of factors.
Accurately estimating biological ages won’t just be a cool trick for AI systems to perform at birthday parties, either. The researchers said that identifying the predictors of aging could help scientists better understand the aging process and reveal keys to healthy aging. Pharmaceutical and insurance companies, as well as medical and longevity industries, would all be keenly interested in that trend.
“Deep biomarkers of aging developed utilizing a variety of data types of aging are rapidly advancing the longevity biotechnology industry. Using biomarkers of aging to improve human health, prevent age-associated diseases and extend healthy life span is now facilitated by the fast-growing capacity of data acquisition, and recent advances in AI. They hold a great potential for changing not only aging research, but healthcare in general,” said Polina Mamoshina, senior scientist at
Insilico Medicine, in the news release.
The researchers predict that AI and aging research will accelerate in the future and note that acceleration is being driven by the increasing number of people entering this field — from universities to corporations — as well as the number of funding sources adding money to those organizations.