Inflammation

What the body actually does, and why it shows up everywhere

Inflammation is not a vague state. It is a coordinated instruction, and following it explains several things on your report that otherwise look unconnected.

Damaged or infected tissue releases signalling molecules locally. One of them, interleukin-6, travels in the blood to the liver, and the liver responds by changing what it manufactures. It is not a general increase in activity. It is a deliberate switch, some proteins up and others down.

Up go C-reactive protein, ferritin, hepcidin and the clotting proteins. Down go albumin, transferrin, and as a consequence of the hepcidin rise, the iron circulating in your blood.

The logic appears to be defensive. Bacteria need iron to multiply, so locking iron away and reducing how much is available in circulation is a way of starving them. The clotting proteins rise because damaged tissue may need sealing. Albumin falls partly because the liver's protein-making capacity has been redirected.

The practical consequence is that one inflammatory event moves markers across 4 different sections of this guide at once. A ferritin that looks healthy, a serum iron that looks low, an albumin that looks low, and a raised CRP can all be a single cold 2 weeks ago rather than 4 separate findings. Reading them as 4 problems is one of the most common ways a panel gets misread.

What the profile looks at

hs-CRP measures C-reactive protein, the protein the liver makes most dramatically in response to the signalling described above, and it can rise many hundredfold. Its job is to mark damaged cells and microbes so the immune system clears them. It rises within hours of a trigger and then halves about every 19 hours once the trigger goes, so the level tracks what is happening now. That combination makes it a good clock.

The hs stands for high sensitivity, and it is not a different substance. It is the same protein measured with an assay able to resolve the low end of the range, which is where the persistent low-grade signal lives. A standard CRP answers whether there is significant inflammation right now. A high-sensitivity CRP can see the background.

ESR, erythrocyte sedimentation rate, is much older and entirely indirect. It measures how fast red blood cells settle in a tube over an hour.

What makes them settle faster explains the whole marker. Red cells carry a surface charge that makes them repel one another. Large proteins produced during inflammation, mainly fibrinogen and immunoglobulins, neutralise that repulsion, so the cells stack together into columns. Heavier columns sink faster. So ESR is not measuring inflammation. It is measuring the consequence of inflammation on how red cells behave.

Two things follow. It moves slowly in both directions, because those proteins are produced and cleared over days rather than hours, and it can stay raised for weeks after the cause has resolved. And because it depends on the cells rather than the proteins alone, anything that changes red cells changes it. That is a direct link to the Blood cells section rather than a quirk of the test.

The two are on different clocks, and that is the point of having both. A raised CRP with a normal ESR suggests something recent. A normal CRP with a raised ESR suggests something that has been running and may now be settling. Neither replaces the other.

The caveat that matters most

hs-CRP is exquisitely non-specific. A very long list of ordinary events will raise it, and the Biomarker Guide covers them. A single raised reading tells you that something provoked a response, not what.

Which is why this marker in particular rewards repetition. One number is noise. The same number across several tests months apart is a pattern.

The cause almost nobody checks

If your hs-CRP sits persistently raised and none of the obvious explanations fit, there is one thing to rule out before going looking for anything exotic. Look in your mouth.

Gum disease is a chronic bacterial infection with a permanent inflammatory response attached, and a very large number of people have it without knowing, because in its early stages it does not hurt. It is not a small signal either. People with the more aggressive form carry CRP levels more than 50% higher than those with the milder form.

And treating it works. Pooled across the randomised trials, treating gum disease lowered CRP by about 0.69 mg/L at 6 months. The authors put that on a par with what lifestyle change or drugs achieve. The biggest falls were in people who started above 3, which is the band that gets flagged as high risk in the first place.

One practical detail if you go and get it sorted. Intensive treatment done in a single long session pushes CRP up first before it comes down. Treatment spread across several visits does not. So if you have just had your gums worked on, leave it a few months before reading anything into an hs-CRP.

A marker of risk rather than a target

This one is easy to get the wrong way round. The evidence that a raised C-reactive protein identifies people at higher cardiovascular risk is strong and consistent. Evidence that lowering C-reactive protein itself reduces that risk, independently of the things that lower it, is not established.

So it earns its place as something that sharpens a risk estimate, not as a number to drive down on its own. The same structural point applies to HDL cholesterol in the lipid section, for the same reason.

But inflammation itself has turned out to be a target

That distinction matters and it is easy to slide past. C-reactive protein is a readout. The inflammation producing it is a process. Nobody has shown that pushing the readout down achieves anything. Several trials have now shown that interfering with the process does.

The first was CANTOS, in 2017. An injected antibody against one of the inflammatory signalling molecules, given to people who had already had a heart attack and whose hs-CRP was 2 or above. It cut cardiovascular events by 15% without touching their cholesterol at all. That was the proof that inflammation is doing something rather than merely marking it. The drug is extremely expensive, was never approved for this, and came with a small but real increase in deaths from infection, so it went nowhere in practice. What it established stayed.

What followed was colchicine, a very old and very cheap anti-inflammatory drug better known for treating gout. Two large trials tested half a milligram a day on top of everything else people were already taking. In those within a month of a heart attack, events fell from 7.1% to 5.5%. In those with long-standing coronary disease, from 9.6% to 6.8% over about two and a half years. European guidance now says low-dose colchicine should be considered in chronic coronary disease.

Three things stop this being as clean as it sounds, and they do not usually get mentioned. More than half the events prevented in the larger trial were procedures to open an artery, rather than heart attacks, strokes or deaths. Deaths from causes other than the heart were higher in the colchicine group, not significantly, but in the wrong direction and unexplained. And the most recent large trial, in people having a heart attack, found no benefit at all.

So the position is this. Inflammation is a real target, there is now a cheap drug aimed at it, and the benefit is genuine but modest and still argued over, in people who already have established heart disease. None of it is a reason for anybody well to take anything. And none of it changes what your own hs-CRP number is. It is a marker. What it marks has turned out to be worth acting on.