Inflammation: The Cascade and How It Resolves

Cut your finger and a chemical chain reaction fires within seconds: membrane fats are cut loose into arachidonic acid, which splits down two arms — COX makes prostaglandins (pain, fever, redness) and 5-LOX makes leukotrienes (which call in neutrophils). Watch the whole cascade run, see hs-CRP climb in the blood, and — the part almost everyone gets wrong — watch inflammation switch itself off using an active resolution programme built from omega-3 fats.

Try this: let the acute injury run once, then hit NSAID and watch prostaglandins collapse while leukotrienes rise — that is the shunt. Then compare ω-3 replete with ω-3 deficient, and finally press Chronic to see what "failure to resolve" actually looks like.

The Science in Plain Language

Inflammation is a repair programme, not the enemy

Almost everything written about inflammation treats it as damage — a fire to be put out, a thing to be “reduced.” That framing is backwards, and it leads people to some genuinely bad decisions.

Inflammation is what your body does when tissue is broken or invaded. It is a controlled demolition-and-rebuild programme. It brings blood, plasma proteins and white cells to the exact place they are needed, kills whatever got in, clears away the wreckage, and then hands the site over to the cells that lay down new tissue. People born unable to mount it — those with severe neutrophil defects, for example — do not enjoy a life free of swelling. They die of infections in childhood.

So the goal is never “no inflammation.” The goal is inflammation that finishes. A cascade that fires hard, does its job, and then shuts itself off completely. The trouble — the arthritis, the atherosclerosis, the metabolic disease — comes almost entirely from cascades that start normally and never finish. That distinction is what this whole page is about.

The five cardinal signs, and the molecule behind each one

The Roman writer Celsus described four signs of inflammation about two thousand years ago, and Galen later added a fifth. They are still the best starting point, because each one now maps onto a specific molecule you can see in the animation.

1. Redness (rubor) and heat (calor) — the small vessels near the injury widen. Histamine from mast cells does this within seconds; prostaglandins keep it going for hours. More warm blood arrives, so the skin looks red and feels hot. Notice in the animation that the vessel physically gets fatter.
2. Swelling (tumor) — the endothelial cells lining the smallest veins pull apart from each other, leaving gaps. Protein-rich fluid leaks out into the tissue. That fluid is not just water; it carries antibodies, complement and clotting factors to the site. The swelling is the delivery.
3. Pain (dolor) — this one surprises people. PGE₂ does not itself cause pain. It sensitises the pain nerve endings, lowering the threshold at which they fire, so that pressure and movement that would normally feel like nothing now feel like injury. Bradykinin and low pH do the actual poking. PGE₂ just turns up the microphone. That is precisely why a drug that only blocks PGE₂ production can be a real painkiller without touching an opioid receptor.
4. Loss of function (functio laesa) — you stop using the part. This is a feature, not a bug. A joint that hurts to move is a joint you rest.

Press NSAID in the animation and watch the four badges on the left. The pain badge dims. Redness, heat and swelling stay lit. That is not a quirk of the model — it is exactly what the pharmacology predicts, because those three are driven by histamine and leukotrienes, which an NSAID does not touch.

Two arms from one fat: COX and 5-LOX

Every cell membrane in your body is built from phospholipids, and buried in those phospholipids is arachidonic acid — a 20-carbon omega-6 fat. It sits there inertly until an enzyme called phospholipase A₂ cuts it free. That single cut is the starting gun for the entire lipid cascade.

Free arachidonic acid then goes down one of two arms:

• The COX arm. Cyclooxygenase — COX-1 (present all the time, housekeeping) and COX-2 (switched on by inflammation) — turns it into prostaglandins and thromboxane. PGE₂ is the star: it widens vessels, sensitises pain nerves, and, when it reaches the hypothalamus, resets your thermostat upward. That is your fever.
• The 5-LOX arm. 5-lipoxygenase turns it into leukotrienes. LTB₄ is one of the most powerful neutrophil-attracting molecules known — it is the flare that calls the infantry. The cysteinyl leukotrienes (LTC₄, LTD₄, LTE₄) constrict airways, which is why leukotriene blockers are asthma drugs.

These two arms compete for the same pool of arachidonic acid. Hold that thought — it is the key to the next section.

Why an NSAID kills the pain but not the fire

Aspirin, ibuprofen, naproxen, diclofenac and celecoxib all do one thing: they block COX. PGE₂ production collapses. The pain nerves stop being sensitised, the thermostat resets, and you feel dramatically better within an hour. This is a real, valuable, well-earned effect, and nothing below is an argument against taking an NSAID when you are in pain.

But watch the side panel while the NSAID is on. hs-CRP barely moves. IL-6 barely moves. NF-κB barely moves. The neutrophils still pour in. You have silenced the alarm bell without touching the fire, because the cytokine arm — the part that actually drives the systemic inflammatory response — runs on a completely separate track from COX.

Two further things happen, and both are worth knowing.

The shunt. If COX is blocked but phospholipase A₂ is still cutting arachidonic acid loose, that substrate has to go somewhere. More of it goes down the 5-LOX arm. Leukotrienes go up. In most people this is a modest, unimportant shift. In the roughly one person in twenty with asthma who has aspirin-exacerbated respiratory disease, it is dangerous: a single aspirin or ibuprofen triggers a surge of cysteinyl leukotrienes and a severe asthma attack. That syndrome is the clearest human proof that the shunt is real.

The resolution problem. This is the subtle one. It turns out that PGE₂ — the same molecule causing your pain — is also the signal that tells neutrophils to switch on 15-lipoxygenase and start making the pro-resolving mediators described below. PGE₂ both starts the fire and lights the fuse that will eventually put it out. Block COX completely and you can blunt that switch, which in animal models delays resolution. The human evidence is thin, and this is emphatically not a reason to suffer through pain. But it is a reason to be sceptical of taking an NSAID every single day for years as a general “anti-inflammatory” strategy. Aspirin is the interesting exception: instead of shutting COX-2 down entirely, aspirin chemically acetylates it, and the acetylated enzyme starts producing a different product — aspirin-triggered lipoxins, which are pro-resolving. Aspirin is the one COX inhibitor that keeps a foot in the resolution door.

Why a corticosteroid is a sledgehammer

Prednisone, dexamethasone and hydrocortisone act much further upstream. They suppress phospholipase A₂ activity, so arachidonic acid is never cut loose in the first place — which means both arms shut down together. Press Corticosteroid and watch both bars collapse at once. Nothing else in the animation does that.

They also do something an NSAID cannot: the activated glucocorticoid receptor goes into the nucleus and directly represses NF-κB, the master transcription switch. TNF-α, IL-1β and IL-6 all fall. IL-6 falls, so the liver stops making CRP, so hs-CRP falls too. This is why steroids work when nothing else will — in a severe asthma attack, a lupus flare, giant cell arteritis, an acute transplant rejection.

The price is exactly proportional to the power. You have not blocked one enzyme; you have turned down the entire inflammatory transcription programme, everywhere in the body, in every tissue. Long-term that buys you infection risk, high blood sugar, bone loss, thin skin, muscle wasting, cataracts, mood disturbance and suppression of your own adrenal glands. A short course is usually a fine trade. A long one is a serious decision with a real bill attached.

One nuance worth knowing, because it complicates the simple story: glucocorticoids also induce annexin A1, a genuinely pro-resolving protein that helps macrophages clear dead neutrophils. Steroids do not just suppress — they push some of the resolution machinery too. That is why, in the animation, the resolution index under a steroid ends up middling rather than terrible.

Resolution is ACTIVE — and this is the part almost everyone gets wrong

For most of the twentieth century, the textbook assumption was that inflammation ends passively. The stimulus goes away, the chemical signals wash out and diffuse away, and the fire simply burns down for lack of fuel. Fade to black.

That is wrong. Work led by Charles Serhan and colleagues from the 1990s onward showed that resolution is its own active, programmed, biochemically distinct phase, with its own dedicated molecules and its own switch. The body does not let the fire burn out. It sends a fire crew.

Those molecules are the specialised pro-resolving mediators, or SPMs:

• Resolvins — the E-series made from EPA, the D-series made from DHA. Both are omega-3 fats from oily fish or algae.
• Protectins and maresins — also from DHA. Maresins come from macrophages, which is where the name comes from (macrophage resolution).
• Lipoxins — made, remarkably, from arachidonic acid itself, the same omega-6 fat that started the whole thing.

Here is what makes them so different from every anti-inflammatory drug we have. An NSAID or a steroid works by subtraction — it removes a signal. SPMs work by addition — they issue new, positive instructions:

1. Stop sending more neutrophils. They shut the door at the vessel wall, without killing the neutrophils already there.
2. Switch the macrophages. They tell macrophages to change phenotype — from the attacking, cytokine-spewing state to a clean-up-and-repair state — and to start eating the dead neutrophils. That eating has a name: efferocytosis. It is the single most important step in ending inflammation, because a dead neutrophil that is not eaten eventually bursts and spills its contents, which are themselves danger signals. Failure to clear the dead is how a fire restarts itself.
3. They do not immunosuppress. This is the crucial difference. SPMs actually enhance bacterial clearance while switching off the inflammation — a combination no steroid can offer.

An honest caveat, because this field attracts more hype than it has earned: most of the SPM evidence is from cells and animals, plus early human work. Supplements sold as “SPM complexes” are not proven therapies for anything, and you should not buy them expecting a drug effect. What is solidly established is the biology — that resolution is active, that it is built from EPA and DHA, and that when it fails, inflammation becomes chronic.

The practical takeaway is simpler than the supplement aisle suggests: the substrate for the D-series and E-series resolvins is EPA and DHA, and the amount of EPA and DHA in your cell membranes is set by what you eat over months. Compare ω-3 replete with ω-3 deficient in the animation. The cascade fires almost identically. What differs is how well, and how completely, it ends.

What hs-CRP actually measures — and what the numbers mean

C-reactive protein is an acute-phase protein. It is made by liver cells, and it is made almost entirely on the orders of one cytokine: IL-6. That is the whole chain, and you can watch it in the animation: NF-κB → IL-6 → travels in the blood to the liver → liver makes CRP → CRP appears in your blood test.

“hs” means high-sensitivity — it is the same protein, measured with an assay precise enough to resolve the low end, which is where cardiovascular risk stratification lives. The kinetics are worth memorising because they explain most of the confusion around the test:

• It starts to rise about 6 hours after the insult — the liver needs time to transcribe and secrete it.
• It peaks at roughly 24 to 48 hours.
• Its half-life is about 19 hours, and that half-life is fixed — it does not change with disease, drugs or anything else. Clearance is constant, so the only thing that determines your CRP level is how much is being made right now. When CRP falls, it means production stopped. That is why serial CRP is such a good way to track whether a treatment is working.

The risk bands used for cardiovascular stratification (the AHA/CDC consensus categories) are:

• Under 1 mg/L — low risk.
• 1 to 3 mg/L — average risk. This is where most of the developed world sits, and where the “chronic low-grade inflammation” story plays out.
• Above 3 mg/L — high relative risk.
• Above 10 mg/L — stop reading it as a cardiovascular number. This is the single most common misinterpretation. A value over 10 means something acute is going on: an infection, an injury, recent surgery, a flare of an autoimmune disease, a bad dental abscess. The correct response is to look for that cause, treat it, and repeat the test a couple of weeks later once you are well. Do not tell someone with a CRP of 40 that they have terrible heart risk. Tell them to find out why it is 40.

hs-CRP is completely non-specific. It tells you that inflammation exists somewhere. It cannot tell you where or why. It is raised by infection, injury, surgery, autoimmune disease, obesity, smoking, poor sleep, oestrogen and hormone therapy, and even gum disease. It should never be used to diagnose anything on its own.

A note on its older cousin, the ESR: the erythrocyte sedimentation rate is driven mostly by fibrinogen and immunoglobulins, and it moves over days-to-weeks rather than hours. CRP rises faster and falls faster. In an acute problem CRP is the more useful test; in something smouldering over months, ESR sometimes tracks better. They are not interchangeable, which is why rheumatologists often order both.

Chronic low-grade inflammation, and what actually moves it

Press Chronic and watch what happens. The cascade fires normally. The neutrophils arrive normally. And then the resolution phase — the part that should end it — never completes. Debris is not fully cleared. The macrophages stay in attack mode. And hs-CRP, instead of returning under 1 mg/L, parks itself somewhere between 1 and 3 mg/L and stays there. For years.

That state has a name: inflammaging. It is not one disease. It is a persistent, low-amplitude inflammatory tone, and it is statistically associated with atherosclerosis, insulin resistance, frailty, depression and dementia. The signal is small — a CRP of 2.4 is one-twentieth of what a bad infection produces — but it never stops, and decades of it appear to matter.

Where does the low-grade signal come from? The honest answer is: several places at once.

• Visceral fat. Fat around the organs is not inert storage; it is an active endocrine tissue packed with macrophages, and it secretes IL-6 directly. A substantial share of the IL-6 circulating in a resting person comes from adipose tissue. This is the single biggest lever for most people.
• Loss of muscle, poor sleep, smoking, chronic periodontal disease, and a leaky or dysbiotic gut barrier all add to the tone.
• Senescent cells — cells that have stopped dividing but refuse to die, and secrete a cocktail of inflammatory signals known as the senescence-associated secretory phenotype (SASP). These accumulate with age.

Now the part that matters, and where most health advice goes wrong. What genuinely lowers hs-CRP:

• Losing visceral fat. Nothing else comes close. This is the highest-yield intervention that exists for chronic low-grade inflammation.
• Regular exercise and keeping muscle. Here is a lovely piece of nuance: a hard workout raises IL-6 sharply — but muscle-derived IL-6 is a completely different signal from fat-derived IL-6. Muscle IL-6 triggers the release of anti-inflammatory mediators like IL-10 and IL-1ra. The same molecule, from a different tissue, in a different context, does the opposite thing. Biology is contextual, and this is one of the cleanest examples.
• Sleep. Restricting sleep raises IL-6 and CRP in controlled experiments. This is not a soft factor.
• Stopping smoking. Reliably lowers CRP.
• A whole-food, high-fibre dietary pattern, and getting your omega-3 intake up (oily fish, or algae if you do not eat fish) so that EPA and DHA are actually present in your membranes when the resolution programme needs them.
• Treating the specific thing. If your CRP is up because of gum disease, rheumatoid arthritis or sleep apnoea, treating that is the intervention.

And the honest negative: a stack of “anti-inflammatory” supplements is not a substitute for any of the above. Curcumin and quercetin have real mechanisms and real signals in small trials, and curcumin in particular has a serious bioavailability problem that most products do not solve. They may be worth a modest place. They will not undo visceral obesity, four hours of sleep, or a sedentary decade, and no honest reading of the evidence says otherwise.

It is worth adding that inflammation as a cause of cardiovascular disease is no longer speculative. Statins lower hs-CRP independently of how much they lower LDL. A trial of canakinumab — an antibody that blocks IL-1β and nothing else, with no effect on cholesterol at all — reduced recurrent heart attacks, which is about as clean a demonstration as medicine gets that the inflammation itself is doing damage. It also caused a small excess of fatal infections, which is the recurring lesson of this entire page: you cannot suppress inflammation without giving something up. Low-dose colchicine has since shown a similar cardiovascular benefit. The trials of high-dose omega-3 are genuinely contested — one large trial of purified EPA showed a clear benefit and another of a mixed EPA/DHA product showed none, and the reasons are still argued over. Anyone who tells you that question is settled is not reading the same literature.

How to read this animation honestly

Every visualisation is a simplification, and you deserve to know exactly where the line is.

Real: the hs-CRP scale, its risk bands (<1 / 1–3 / >3, and >10 meaning “find the acute cause”), and its kinetics — the ~6-hour lag, the 24–48-hour peak, and the fixed ~19-hour half-life. The mechanism is real too: the enzymes, the mediators, the order of events, the drug targets, and the direction and rough size of every effect you can trigger with the buttons.

Illustrative: every other number in the side panel. The neutrophil count per field, the cytokine concentrations, the eicosanoid bars as a percentage of an untreated peak, and above all the “resolution index” — which is not a lab test and does not exist. It is a teaching device that combines SPM tone, macrophage phenotype and how much of the debris has actually been cleared, so that you can see a thing that is otherwise invisible: whether this inflammatory episode is going to finish, or not.

Real inflammation is also messier than this. Macrophages are not two boxes labelled M1 and M2; they sit on a continuous spectrum and change minute to minute. Mast cells, platelets, complement, the clotting cascade, T cells and the nervous system are all involved and are not drawn here. What is drawn is the spine of the thing — and the spine is enough to understand why an NSAID relieves your pain but does not lower your CRP, why a steroid is both a miracle and a mortgage, and why the most interesting question in inflammation is not how it starts, but whether it ever properly stops.

↑ Back to the animation

Connections

• All Interactive Visualizations
• The Immune Response (visualization)
• Gut Barrier & Microbiome (visualization)
• hs-CRP Test
• Inflammatory Markers
• ESR Test
• Omega-3 Index
• Omega-3 Fatty Acids
• Anti-Inflammatory Diet
• Aspirin
• Curcumin
• Quercetin
• Rheumatoid Arthritis