Chronic inflammation has been named as a driver of heart disease for the better part of three decades. That claim now has its cleanest test — and the answer is not the one the field expected.
In July 2026, a large phase 3 trial reported that a drug which switched off one of the body's central inflammatory signals delivered no cardiovascular benefit whatsoever. The pharmaceutical company behind it shed roughly $30.7 billion in market value in a single trading day, falling from $229.2 billion to $198.5 billion [1].
The result is worth understanding properly, because it changes what "lowering inflammation" should mean in practice — and it does not mean what most supplement marketing implies.
Table of Contents
- What is IL-6, and what does it do?
- Why did researchers expect blocking IL-6 to prevent heart attacks?
- Has any anti-inflammatory drug ever reduced heart attacks?
- What did the ZEUS trial find?
- Why didn't blocking IL-6 work?
- Does inflammation still matter for heart disease?
- How do you lower inflammation at its source?
- References
What is IL-6, and what does it do?
Interleukin-6 — usually shortened to IL-6 — is a signalling protein released by immune cells. Its job is to raise the alarm: after an injury, an infection, or a period of physical stress, IL-6 levels climb and the body's repair and defence systems switch on. C-reactive protein, or CRP, is produced downstream of that signal, which is why a CRP blood test is used as a rough proxy for how much inflammation is present.
What made IL-6 interesting to researchers was not its role in acute illness but its behaviour with age. A 1993 report described something unexpected in routine blood work: young, healthy people have almost no detectable IL-6 in circulation, while older people frequently do — with no infection and no obvious illness to account for it. The author's description of IL-6 as a cytokine of interest to gerontologists stuck [2].
By 2000 that observation had a name. Researchers proposed that ageing itself involves "a progressive increase in proinflammatory status" — a permanently raised, low-grade inflammatory background that they called inflammaging [3]. Under this model IL-6 was not simply a marker of getting older. It was a candidate cause.
Population data appeared to support that reading. Pooling the available studies, people in the top band of circulating IL-6 die at roughly a fifty percent higher rate from any cause than people in the bottom band [4]. They also lose muscle strength sooner and lose functional independence earlier. A 2006 review pulled the field together and set out the stakes plainly: if raised IL-6 turns out to be causally responsible for physical decline in later life, "rather than merely reflecting other processes", then interfering with it could be a major breakthrough [5].
Why did researchers expect blocking IL-6 to prevent heart attacks?
Observational associations are weak evidence on their own — sick people are inflamed, so inflammation and bad outcomes travel together without either one causing the other. What moved IL-6 from interesting to compelling was human genetics.
The technique is Mendelian randomisation. Gene variants are allocated at conception, effectively at random and long before lifestyle, diet or illness can interfere. Some people inherit a variant that mildly dampens IL-6 signalling for their entire life. Comparing those people with everyone else approximates a randomised trial that began at birth and from which nobody could withdraw.
In 2012 a consortium ran exactly that analysis across roughly 133,000 participants. Carrying the IL-6-dampening variant was associated with lower rates of coronary heart disease, and the statistical signal was strong — a p-value with four zeros after the decimal point before the first significant digit [6].
A later analysis extended the same approach to how long people live, reporting that carriers of the protective variant tended to have longer-lived parents — both mothers and fathers [7]. The cardiovascular signal and the ageing signal pointed in the same direction.
This mattered because the identical method had already proven itself elsewhere. Applied to LDL cholesterol and ApoB, Mendelian randomisation agrees closely with what the drug trials found, and shows that lifelong exposure to lower LDL delivers substantially more risk reduction than starting a statin in middle age [8]. A method with that track record pointing at IL-6 was taken seriously.
Has any anti-inflammatory drug ever reduced heart attacks?
Yes — once, convincingly. The CANTOS trial, reported in 2017, enrolled 10,061 people who had already survived a heart attack and still had a raised CRP. They received canakinumab, an antibody against interleukin-1 beta, which sits one step upstream of IL-6 in the inflammatory cascade. Shut down the upstream switch and the signals below it quieten.
Major cardiovascular events fell, with a hazard ratio of 0.85 — for every hundred events in the placebo arm, about eighty-five occurred on the drug. Crucially, cholesterol was not altered at all, so the benefit could not be attributed to lipid lowering [9]. It was the first time anyone had separated the two.
The same trial carried a warning. Fatal infection or sepsis occurred at 0.31 per 100 person-years on the drug versus 0.18 on placebo [9] — a signal that becomes important later in this story.
A cheaper approach was tested next. CIRT gave low-dose methotrexate to nearly five thousand participants and found no cardiovascular benefit at all. But the investigators then checked whether the drug had actually done anything to inflammation, and it had not: "methotrexate did not result in lower interleukin-1β, interleukin-6, or C-reactive protein levels than placebo" [10]. A drug that never engages its target cannot test the hypothesis. The inflammation theory survived intact, and the obvious next step was to hit IL-6 directly.
What did the ZEUS trial find?
ZEUS tested ziltivekimab, a monoclonal antibody directed at IL-6 itself. The design was close to ideal for the question. It randomised 6,376 participants, every one of whom had documented residual inflammation on a CRP test. All were already on cholesterol-lowering treatment, with a mean LDL of 77.7 mg/dL — treated, though not down to the level current practice would aim for [11]. Whatever cardiovascular risk remained in this group, a meaningful share of it was inflammatory rather than lipid-driven.
Pharmacologically, the drug performed. Free IL-6 fell. CRP fell. Target engagement was not in question.
The cardiovascular result was flat. The hazard ratio for major adverse cardiovascular events came in at 0.99, with a confidence interval running from 0.88 to 1.11 [12]. That is not a small benefit that missed significance. It is no signal at all.
The safety profile echoed CANTOS. A higher proportion of participants on ziltivekimab developed serious infections than on placebo, which is exactly what suppressing IL-6 would be predicted to do. There was no difference between the groups in deaths from any cause [12].
Why didn't blocking IL-6 work?
The most useful way to read ZEUS is not that inflammation is irrelevant to heart disease, but that IL-6 was the wrong thing to remove.
IL-6 and CRP are messengers. Their function is to mobilise the immune system against infection and to coordinate tissue repair. Deleting the message does not remove whatever is generating it — and it does remove the message's legitimate uses. That both CANTOS and ZEUS produced more serious infections is not a surprising side effect; it is the predictable consequence of switching off a defence signal.
Infection defence is not the only casualty. A 2019 randomised controlled trial recruited people carrying substantial abdominal fat, which is itself a significant source of circulating IL-6. Participants were randomised on two axes simultaneously: twelve weeks of cycling training or none, and an IL-6 blocking antibody or placebo. Training reduced visceral fat as expected — in every group except the one receiving the IL-6 blocker. There, the same twelve weeks of exercise produced no visceral fat loss.
The authors concluded that "IL-6 is required for exercise to reduce visceral adipose tissue mass" [13].
The molecule cast as the villain turns out to be part of the machinery through which exercise delivers its benefits. That is a strong argument against treating a raised inflammatory marker as something to be abolished.
Does inflammation still matter for heart disease?
It does. Nothing in ZEUS overturns the observational and genetic evidence linking inflammatory burden to cardiovascular risk. What the trial demonstrates is narrower and more practical: suppressing the signal, on its own, does not reduce events.
The distinction is between a cause and a readout. A raised CRP in someone with excess visceral fat, poor glucose control or untreated sleep apnoea is reporting a real problem. Silencing the report leaves the problem in place. Addressing whatever is driving the inflammation brings the marker down as a consequence — and, unlike an antibody, does not compromise immune defence or blunt the response to training.
How do you lower inflammation at its source?
The most common driver in general practice is visceral fat — the fat packed around the abdominal organs, which behaves as an active endocrine tissue and secretes IL-6 continuously.
When patients lose that fat, IL-6 and CRP typically fall with it, because the source has been removed rather than the signal muted.
GLP-1 medications are notably effective at shifting this specific fat depot. In a post hoc analysis of one trial, conducted in Japanese adults, abdominal visceral fat area fell by forty percent on semaglutide against seven percent on placebo [14]. Inflammatory markers track that change: across the semaglutide weight-loss trials, CRP fell by around forty percent [15].
Where residual inflammation persists after weight has been addressed, low-dose colchicine is an inexpensive option with randomised evidence behind it. A meta-analysis of more than 21,800 patients found colchicine reduced cardiovascular events by about 25 percent, although the authors graded the variation between the constituent trials as high [16].
Beyond that, the levers that reduce cardiovascular risk are the familiar ones and remain the priority: visceral fat, insulin resistance, blood pressure, psychological stress, smoking, and ApoB. ZEUS is a $30 billion demonstration that shortcuts around them do not work.
References
2. https://pubmed.ncbi.nlm.nih.gov/8426042/
3. https://pubmed.ncbi.nlm.nih.gov/10911963/
4. https://pubmed.ncbi.nlm.nih.gov/28866452/
5. https://pmc.ncbi.nlm.nih.gov/articles/PMC2645627/
6. https://pubmed.ncbi.nlm.nih.gov/22421340/
7. https://doi.org/10.1038/s41525-019-0097-4
8. https://pubmed.ncbi.nlm.nih.gov/23083789/
9. https://pubmed.ncbi.nlm.nih.gov/28845751/
10. https://pubmed.ncbi.nlm.nih.gov/30415610/
11. https://pubmed.ncbi.nlm.nih.gov/41369941/
13. https://pubmed.ncbi.nlm.nih.gov/30595477/
14. https://pubmed.ncbi.nlm.nih.gov/40189961/
