Inflammatory markers, and how they fit together

Inflammation in a blood test is measured indirectly. Nobody can draw "inflammation". What labs measure is a set of molecules whose levels change when the immune system is active. Understanding which is which explains why they disagree with each other, why they spike after a cold, and why only one of them has strong evidence for use in healthy people.

The acute-phase response

When tissue is injured or infected, immune cells release signalling proteins called cytokines. The best known are interleukin-6 (IL-6), interleukin-1 and tumour necrosis factor alpha (TNF-alpha). These travel to the liver and change what it makes. As Gabay and Kushner described in their 1999 New England Journal of Medicine review, the liver ramps up "positive" acute-phase proteins, including C-reactive protein, fibrinogen and ferritin, and dials down "negative" ones, most importantly albumin. IL-6 is the main driver of CRP production. So the markers sit at different points of one chain: cytokines upstream, liver proteins downstream.

hs-CRP: the one with outcome data

CRP has been measured for decades to detect infection, where values run in the tens or hundreds of mg/L. The high-sensitivity assay measures down to well below 1 mg/L, which is the range relevant to low-grade inflammation in healthy people. The 2003 CDC and American Heart Association statement set the categories still used: under 1 mg/L lower risk, 1 to 3 mg/L average, above 3 mg/L higher. It recommended two measurements, averaged, ideally two weeks apart in someone metabolically stable, and treating a value above 10 mg/L as a sign of acute inflammation to be repeated rather than as cardiovascular risk. The 2018 ACC/AHA cholesterol guideline lists hs-CRP of 2 mg/L or more as a risk-enhancing factor.

Two randomised trials explain why hs-CRP is taken seriously.

JUPITER showed that people selected by hs-CRP benefit from a statin, though a statin lowers LDL too, so it does not isolate inflammation. CANTOS is the cleaner test: it lowered hs-CRP without touching lipids, and events fell. It was a trial in people who had already had a heart attack, so it does not translate directly to healthy people, and canakinumab is not used for this purpose.

ESR and fibrinogen

The erythrocyte sedimentation rate measures how far red cells fall in a tube in an hour. Fibrinogen and other proteins make red cells clump and fall faster, so ESR rises with inflammation, slowly and non-specifically. It also rises with age and runs higher in women. A 1983 BMJ paper proposed the rule of thumb still taught: the upper limit is about age divided by two in men and (age plus 10) divided by two in women. Fibrinogen itself is a clotting protein and an acute-phase protein, so it rises with inflammation and with smoking. The ESR and fibrinogen page has the detail.

Ferritin and albumin

These two are the reason inflammation reading needs care. Ferritin is the body's iron store, so a low value reliably means low iron. But ferritin is also a positive acute-phase protein, so a high value can mean iron overload, liver disease, or inflammation, and a "normal" ferritin in someone with inflammation can hide iron deficiency. Albumin goes the other way, falling during inflammation. A low albumin can be inflammation, liver disease, kidney loss or poor nutrition. Both are read alongside hs-CRP for that reason.

IL-6 and TNF-alpha

These are the signals rather than the response. They change faster than CRP, they are influenced by exercise and time of day, and assays from different manufacturers give different numbers. No major guideline uses them for routine risk assessment in healthy people. They are useful in research and in some specific clinical settings, and they make most sense as a trend at one lab rather than as a single value against a range.

Homocysteine

Homocysteine is an amino acid intermediate, not an acute-phase protein. It rises with low vitamin B12 or folate, with reduced kidney function, and with some genetic variants. It is associated with cardiovascular disease in observational studies, but as the Cochrane review above shows, lowering it with B vitamins did not reduce heart attacks. It is often grouped with inflammatory markers because it is read in the same cardiovascular context.

Reading them together

A single raised marker after a cold, a hard training week or a dental procedure is the commonest finding and usually means nothing chronic. hs-CRP in particular varies a lot within the same person: a 2024 meta-analysis in PLOS ONE found a median within-person coefficient of variation of about 44 percent. That is why repeat measurement matters more here than for almost anything else; see biological variation and retesting. On a GLP-1, hs-CRP tends to fall with weight; the GLP-1 blood work page has the trial figures.

In the app

Peptide AI's inflammatory panel groups all eight markers above, scores them together, and plots each against your dose timeline, with notes from the training log and illness days alongside so a spike has context. See blood panels in the app and bloodwork tracking.