Inflammatory Age: The Clock That Scored a 105-Year-Old at 25

The centenarian who broke the scale

Among the blood samples that built Stanford's inflammatory clock were those of a group of Italians from Bologna who had done something statistically improbable: they had lived past 100. Twenty-nine exceptionally long-lived individuals, weighed against a comparison group of eighteen adults aged 50 to 79.^[1]

The centenarians did not merely score well. Their inflammatory ages came in an average of roughly forty years below their chronological ages, and one man of 105 returned a figure of 25. Not 25 years younger — 25 years old. By the only measure that mattered to the algorithm, his immune system was that of a graduate student.

It is a wonderful anecdote, and it has been doing energetic work in the marketing of longevity clinics ever since. Cytokine panels now sit on premium diagnostic menus beside the epigenetic clocks, priced accordingly. The implicit promise is elegant: here is the fire that ages you, here is your score, and here — for a fee — is the protocol to bring it down.

The promise deserves more scrutiny than it usually gets, and the most damaging evidence against it comes not from sceptics but from the field's own founders.

What the inflammatory clock actually measures

The instrument itself is serious work. Published in Nature Aging in 2021 out of the Stanford 1000 Immunomes Project, iAge was derived by applying deep learning to the circulating immune proteins of 1,001 people spanning ages 8 to 96.^[1] From that noise the model extracted a single number that tracked four things clinicians genuinely care about: multimorbidity, immunosenescence, frailty and cardiovascular aging.

The logic behind it is the concept of inflammaging — chronic, low-grade, systemic inflammation that rises with age in the absence of any infection to justify it. Unlike the acute inflammation that follows a splinter or a virus, inflammaging has no off switch and no obvious provocation. It is now routinely described as intertwined with immunosenescence, the progressive decline of immune competence, and implicated in cancer, dementia and neurodegeneration.^[5] Much of the tissue-level fuel is thought to come from senescent cells, which is why the case for clearing zombie cells and the case for lowering inflammation are, mechanistically, close to the same argument.

What lifts the iAge paper above correlation is what its authors did next. Having identified CXCL9 — an interferon-induced chemokine that recruits T cells — as the strongest single contributor to the score, they went looking for causation. CXCL9 turned out to be involved in cardiac aging, adverse cardiac remodelling and poor vascular function. Then, in aging human and mouse endothelial cells showing loss of function, cellular senescence and the hallmark phenotypes of arterial stiffness, they silenced it. The phenotypes reversed.^[1]

That is a real mechanistic finding, not a dashboard metric. It is entirely reasonable to conclude that chronic inflammatory signalling does damage, and that CXCL9 is one of the hands doing it.

The 2025 result that reframes the whole field

Here the story turns, and it turns in Nature Aging again.

In 2025, a large international group led from Sherbrooke and Columbia asked a question that had somehow gone unasked for two decades: is inflammaging actually universal?^[2] They took a 19-cytokine inflammaging axis identified in the Italian InCHIANTI cohort and tested whether it reproduced elsewhere. In the Singapore Longitudinal Aging Study — another industrialised population — it broadly did, differing only in IL-6 and IL-1RA.

Then they took it to the Tsimane, forager-horticulturalists of the Bolivian Amazon, and to the Orang Asli of Peninsular Malaysia. In both, the cytokine axis had a markedly different structure. It showed little to no association with age. And it showed no association with age-related disease. The authors' conclusion is unusually blunt for a journal of that standing: inflammaging, measured this way, "appears to be largely a byproduct of industrialized lifestyles, with major variation across environments and populations."^[2]

The detail that gives this its weight sits in the author list. Claudio Franceschi — the Bologna gerontologist who coined the term inflammaging in 2000, and who also appears on the iAge paper — is a co-author. This is not an outside attack on the concept. It is the concept's own author, and much of its establishment, publishing its limits.

The Tsimane paradox: inflamed blood, immaculate arteries

To see why this matters, look at what the Tsimane actually have in their blood, and what it does to them.

In 2017 a team including several of the same anthropologists ran CT calcium scoring on 705 Tsimane adults aged 40 and over and compared them against 6,814 participants in the Multi-Ethnic Study of Atherosclerosis.^[3] The results are among the most striking in modern cardiology. Of the 705, 596 — 85% — had no coronary artery calcium whatsoever. Among those over 75, 65% still had a score of zero, and only 8% had a score above 100: a five-fold lower prevalence of significant disease than industrialised populations at every age band tested.

The paper's own summary is that the Tsimane have "the lowest reported levels of coronary artery disease of any population recorded to date."

Now the part that should stop any longevity investor mid-sentence. In that same population, high-sensitivity C-reactive protein was elevated beyond the standard clinical cutoff in 360 participants — 51% of the cohort.^[3] They carry a heavy lifelong burden of infection and parasites, and their inflammatory markers reflect it faithfully.

So: half the population walking around with inflammation that would trigger a serious conversation in any Western longevity clinic, and simultaneously the cleanest coronary arteries ever documented in a human group. Inflammation and cardiovascular aging, which the inflammatory clock treats as tightly coupled, have here come completely apart.

What happened when medicine drugged inflammation directly

There is one more piece of evidence, and it is the most expensive ever assembled on this question.

The CANTOS trial randomised 10,061 patients with a previous heart attack and CRP of at least 2 mg/L to canakinumab, a monoclonal antibody against interleukin-1β, or placebo.^[4] This was the inflammatory hypothesis of atherosclerosis put on trial with real money. The drug worked as designed: CRP fell by 37 percentage points more than placebo at the 150 mg dose, with no change in lipids at all.

And the cardiovascular benefit was real, but modest. At the 150 mg dose the hazard ratio for the primary endpoint was 0.85 (95% CI 0.74–0.98, P=0.021). Neither the 50 mg nor the 300 mg dose met the prespecified threshold for significance — a non-monotonic dose response that ought to induce caution in anyone.

Two findings from CANTOS rarely make it into wellness summaries. Canakinumab was associated with a higher incidence of fatal infection than placebo. And across all doses there was no significant difference in all-cause mortality: hazard ratio 0.94 (95% CI 0.83–1.06, P=0.31).^[4]

Suppressing inflammation, in other words, bought fewer heart attacks and paid for them in infections, netting out at no measurable change in whether you died. Inflammation is not simply damage to be minimised. It is the immune system doing its job, and the job is not optional.

So can you lower your inflammatory age?

Probably, in the sense that the number will move. Whether moving the number moves your life is a different question, and no trial has yet answered it: nothing has been shown to lower iAge specifically and thereby extend healthy life. Any protocol sold on that particular promise is ahead of its evidence.

What the Tsimane data does support is an answer so unglamorous it is almost impolite. The authors of that paper attribute the result to a lifetime of very low LDL cholesterol, low blood pressure, low blood glucose, normal body-mass index, no smoking, and a great deal of physical activity — noting candidly that the relative contribution of each remains undetermined.^[3] Nobody there is taking a supplement. There is no panel, no membership, no protocol.

The plausible levers on chronic inflammation in an industrialised life are the same ones that appear in every honest longevity conversation: sustained aerobic and resistance training, sleep sufficient to permit overnight immune regulation, visceral fat kept low, and the condition of the gut microbiome, whose age-related drift is one of the more credible proposed sources of the inflammatory background. Interventions that act on immune aging more directly — spermidine and autophagy among them — are interesting and early, and should be held to the same standard as everything else.

The honesty clause

Nothing above shows the inflammatory clock is wrong. Within the industrialised populations where it was built and validated, it predicts frailty, multimorbidity and cardiovascular aging, and the CXCL9 mechanism is real. If you live in a city and eat from a supermarket, your inflammatory age is telling you something true.

What it is not is a measurement of aging itself. It is a measurement of a particular pattern of immune dysregulation that our environment produces and that, in our environment, happens to track decline. Take the same assay somewhere else and the correlation dissolves — which is the signature of a marker of context rather than a law of biology.

Three further limits deserve stating. The centenarian comparison rests on small numbers: 29 long-lived individuals against 18 controls is a striking finding, not a settled one. All of it is observational, and the healthiest people in any cohort are healthy for reasons that no cytokine panel can disentangle. And CANTOS is a standing rebuke to the assumption that pushing a marker down is the same as pushing a disease down — the arrow from marker to outcome bent, in that trial, only slightly and at a price.

The royal verdict

The best reason to pay attention to your inflammatory age is not the number. It is what the number's failure elsewhere reveals.

An assay that predicts decline in Bologna and Singapore, and predicts nothing at all in the Amazon, is not reading a clock installed in the human body at birth. It is reading the accumulated residue of how a particular kind of life is lived — sedentary, over-fed, under-slept, metabolically strained. That is a considerably more useful thing to know than a score, because a clock is something you watch and a residue is something you can, in principle, stop producing.

There is an irony worth sitting with. The Tsimane, whose blood would alarm any concierge physician, have arteries that no amount of money has yet bought anyone in London or Los Angeles. They achieved it without a single measurement. The instinct to quantify is a fine one, and we would not argue against knowing your numbers — but it is worth remembering that measurement is a proxy for a life, and the proxy is not the thing. Whatever else the inflammatory clock has established, it has established that the real constraints on a long life are less exotic, and less purchasable, than the industry built around measuring them would prefer.

Common questions