Investigators at Stanford University School of Medicine and the Buck Institute for Research on Aging have harnessed artificial intelligence to build an inflammatory-aging clock—iAge— that they suggest is more accurate than the number of candles on your birthday cake in predicting how strong your immune system is, how soon you’ll become frail, or whether you have—as yet unseen—cardiovascular problems that could in the future become clinically relevant.
Through their work with the 1000 Immunomes Project, and other cohorts, the scientists also identified a chemokine associated with cardiac aging, which they suggest could be used for the early detection of age-related pathology, and could also represent a potentially modifiable target for intervention.
“Standard immune metrics which can be used to identify individuals most at risk for developing single or even multiple chronic diseases of aging have been sorely lacking,” said David Furman, PhD, Buck Institute associate professor and director of the Artificial Intelligence Platform, director of the 1001 Immunomes Project at Stanford University School of Medicine, a visiting scholar at Stanford’s Institute for Immunity, Transplantation and Infection, and senior author of the team’s study, which is published in Nature Aging. “Bringing biology to our completely unbiased approach allowed us to identify a number of metrics, including a small immune protein which is involved in age-related systemic chronic inflammation and cardiac aging. We now have means of detecting dysfunction and a pathway to intervention before full-blown pathology occurs.”
The team published its findings in a paper titled, “An inflammatory aging clock (iAge) based on deep learning tracks multimorbidity, immunosenescence, frailty and cardiovascular aging.” Lead authors of the study are Nazish Sayed, MD, PhD, assistant professor of vascular surgery at Stanford, and Yingxiang Huang, PhD, senior data scientist at the Buck Institute.
“Every year, the calendar tells us we’re a year older,” Furman explained. “But not all humans age biologically at the same rate. You see this in the clinic—some older people are extremely disease-prone, while others are the picture of health.” This divergence, Furman said, traces in large part to differing rates at which people’s immune systems decline. The immune system, which has evolved to deal with threats such as injuries or infection by pathogens, excels at mounting a quick, intense, localized, short-term, resist-and-repair response called acute inflammation. This positive inflammatory response typically does its job, then wanes within days. But, as the authors noted, while the important role of the immune system in the maintenance of human health and protection against infections has been recognized for over a hundred years, “ … it was only in the past few decades that it has become apparent that inflammatory components of the immune system are often chronically elevated in aged individuals and associated with an increased incidence of cancer, cardiovascular disease, neurodegenerative disorders, and others.” And these observations have led to the concept that inflammation plays a critical role in regulating physiological aging.
So, it seems, as we grow older, a low-grade, constant, bodywide “bad inflammation” begins to kick in. This systemic and chronic inflammation causes organ damage and promotes vulnerability to autoimmunity, as well as to a range of diseases including cancer, heart attacks, strokes, and neurodegeneration. “Contrary to the acute response, which is typically triggered by infection, chronic and systemic inflammation is thought to be triggered by physical, chemical, or metabolic stimuli (‘sterile’ agents) such as those released by damaged cells and environmental insults, generally termed ‘damage-associated molecular patterns’ (DAMPs),” the team explained. This type of inflammation is associated with aging and is characterized by being low-grade and persistent, ultimately leading to collateral damage to tissues and organs.
Researchers studying aging previously identified nine “hallmarks” of the aging process. “… the well-established nine hallmarks of aging, (1) genomic instability, (2) shortening telomere length, (3) epigenetic modifications, (4) loss of proteostasis, (5) deregulated nutrient sensing, (6) mitochondrial dysfunction, (7) cellular senescence, (8) stem cell exhaustion, and (9) altered intracellular communication, have all been shown to be linked to sustained systemic inflammation,” the team noted. However, age-related immune system dysfunction was not part of the mix.
To date, there have been no metrics for accurately assessing individuals’ inflammatory status in a way that could predict future clinical problems and point to ways of addressing them or staving them off, Furman said. But the new study, he claimed, has produced a single-number quantitative measure that may do just that.
For the 1000 Immunomes Project, blood samples were drawn from 1,001 healthy people, aged 8–96, between 2009 and 2016. The samples were subjected to a range of analytical procedures to determining levels of immune-signaling cytokines, the activation status of numerous immune cell types in response to various stimuli, and the overall activity levels of thousands of genes in each of those cells.
For the newly reported studies, the researchers harnessed artificial intelligence to extract from all this data a composite, to generate an effective inflammatory clock, iAge. The strongest predictors of inflammatory age, they found, were a set of about 50 immune-signaling cytokines. Levels of these cytokines, condensed using a complex algorithm, could be used to generate a single-number inflammatory score that tracked well with a person’s immunological response and the likelihood of incurring any of a variety of aging-related diseases.
Back in 2017, the scientists assessed nearly 30 1000 Immunomes Project participants aged 65 or older whose blood had been drawn in 2010. They measured the participants’ speed at getting up from a chair and walking a fixed distance and, through a questionnaire, their ability to live independently (“Can you walk by yourself?” “Do you need help getting dressed?”). Inflammatory age proved superior to chronological age in predicting frailty seven years later.
Furman and his colleagues obtained blood samples from an ongoing study of exceptionally long-lived people in Bologna, Italy, and compared the inflammatory ages of 29 such people (all but one a centenarian) with those of 18 50–79-year-olds. They found that these older people had inflammatory ages averaging 40 years less than their calendar age. One, a 105-year-old man, had an inflammatory age of 25. Furman said that when it comes to health and longevity, the “age” of one’s immune system most certainly trumps the chronological information that can be derived from a driver’s license. “On average, centenarians have an immune age that is 40 years younger than what is considered ‘normal’ and we have one outlier, a super-healthy 105-year-old man (who lives in Italy) who has the immune system of a 25-year-old,” he said.
To further assess inflammatory age’s effect on mortality, Furman’s team turned to the Framingham study, which has been tracking health outcomes in thousands of individuals since 1948. The Framingham study lacked sufficient data on blood-borne protein levels, but the genes whose activity levels largely dictate the production of the inflammatory clock’s cytokines are well known. The researchers measured those cytokine-encoding genes’ activity levels in Framingham subjects’ cells. This proxy for cytokine levels significantly correlated with all-cause mortality among the Framingham participants.
The scientists observed that blood levels of one substance, CXCL9, contributed more powerfully than any other clock component to the inflammatory-age score. They found that levels of CXCL9, a cytokine secreted by certain immune cells to attract other immune cells to a site of an infection, begin to rise precipitously after age 60, on average.
Among a new cohort of 97 25- to 90-year-old individuals selected from the 1000 Immunomes Project for their apparently excellent health, with no signs of any disease, the investigators looked for subtle signs of cardiovascular deterioration. Using a sensitive test of arterial stiffness, which conveys heightened risk for strokes, heart attacks, and kidney failure, they tied high inflammatory-age scores—and high CXCL9 levels—to unexpected arterial stiffness and another portent of untoward cardiac consequences: excessive thickness of the wall of the heart’s left ventricle. “These people are all healthy according to all available lab tests and clinical assessments, but by using iAge we were able to predict who is likely to suffer from left ventricular hypertrophy (an enlargement and thickening of the walls of the heart’s main pumping chamber) and vascular dysfunction,” he stated.
CXCL9 has previously been implicated in cardiovascular disease. A series of in vitro experiments showed that CXCL9 is secreted not only by immune cells, but by endothelial cells—the main components of blood-vessel walls. The researchers’ results indicated that advanced age correlates with a significant increase in endothelial cells’ CXCL9 levels and diminishes endothelial cells’ ability to form microvascular networks, to dilate and to contract.
But through laboratory experiments conducted on tissue from mice and on human cells, the team showed that reducing CXCL9 levels restored youthful endothelial-cell function, suggesting that CXCL9 directly contributes to those cells’ dysfunction and that inhibiting it could prove effective in reducing susceptible individuals’ risk of cardiovascular disease. “Our inflammatory aging clock’s ability to detect subclinical accelerated cardiovascular aging hints at its potential clinical impact,” Furman said. “All disorders are treated best when they’re treated early.”
The authors concluded, “… by applying artificial intelligence methods to deep immune monitoring of human blood we generate an inflammatory clock of aging, which can be used as a companion diagnostic to inform physicians about patient’s inflammatory burden and overall health status, especially in those with chronic diseases … our immune metric for human health can identify within healthy older adults with no clinical or laboratory evidence of cardiovascular disease, those at risk for early cardiovascular aging.”
Furman says the tool can be used to track someone’s risk of developing multiple chronic diseases by assessing the cumulative physiological damage to their immune system. For example, age-related frailty can be predicted by comparing biological immune metrics with information about how long it takes someone to stand up from a chair and walk a certain distance as well as their degree of autonomy and independence. “Using iAge it’s possible to predict seven years in advance who is going to become frail,” he said. “That leaves us lots of room for interventions.”
The authors concluded, “We used artificial intelligence to create a compact representation of these biomarkers and derived an ‘inflammatory clock’ of aging, which takes into account the nonlinear relationship and redundancy of the cytokine network,” they wrote. “This metric tracked with multiple aging phenotypes in the general population and thus, has strong potential for translational medicine, as it could be used as a diagnostic tool for identifying those at risk for both noncommunicable and infectious diseases.”