Albumin

What it actually measures

Albumin is the most abundant protein in your blood, made entirely by your liver, and it does two jobs that look unrelated.

The first is holding fluid inside your blood vessels. Albumin exerts osmotic pressure, essentially pulling water inward, which is why very low albumin produces swelling as fluid leaks into tissue.

The second is transport. Albumin is the general-purpose carrier of your bloodstream, binding and ferrying hormones, fatty acids, calcium, bilirubin and a great many medications.

That second job has a consequence people rarely connect. Because so many drugs travel bound to albumin, a low albumin means more free drug circulating, which changes how medications behave.

Albumin appears on nearly every comprehensive metabolic panel, which means most people reading this have had it measured repeatedly.

And it is very commonly interpreted as a nutritional marker, which is where this page has something worth saying.

Why it is a poor nutritional marker most of the time

The reasoning behind treating albumin as a nutrition marker is intuitive. It is a protein, your liver builds it from amino acids, so inadequate protein intake should lower it.

That reasoning holds in a specific and narrow situation, and fails outside it.

A 2015 review and meta-analysis in Maturitas examined albumin and health in older people and stated the position clearly: albumin is a good nutritional marker only in clinically stable people, because inflammation, notably through IL-6 and TNF-alpha, is a main driver of low albumin.

The mechanism is that albumin is a negative acute phase reactant. During an inflammatory response the liver shifts production toward acute phase proteins such as C-reactive protein and away from albumin. So albumin falls because inflammation is present, not because protein intake is low.

Notice that this is the mirror image of ferritin, which rises with inflammation. Same underlying process, opposite direction, same interpretive trap.

The same review made a second point worth having: age itself is not a cause of low albumin. So a low value in an older person is a finding rather than an expectation.

What albumin is genuinely good at is prognosis. The review found low albumin independently predicts mortality in older adults across community, hospital and institutional settings, and noted that albumin below 38 g/L in elderly hip-fracture patients is associated with higher risk of post-surgical complications, particularly infections.

That is a real and useful signal. It is a signal about overall illness burden and inflammation rather than about what someone has been eating.

What moves it

Down: inflammation of any kind, which is the most common reason and the one most often misread. Liver disease, since the liver makes it. Kidney disease with protein loss in urine, which is why urine protein is checked alongside. Malabsorption and inflammatory bowel disease. Genuine severe protein malnutrition, which does lower it but is far from the usual explanation. Overhydration, through dilution.

Up: dehydration, which concentrates the blood and raises it without changing the actual amount. Prolonged tourniquet time during the draw, which does the same thing locally. Genuinely high albumin is uncommon and is usually a dilution artefact rather than a finding.

The honest hedge

I am not a doctor and I am not diagnosing anyone.

The correction I most want to offer here is about what a low albumin means. It is frequently read as eat more protein, and that is usually the wrong conclusion.

In someone with any inflammatory process, and that includes infection, injury, autoimmune disease, significant obesity and a long list of others, albumin falls for reasons that protein intake will not fix. Adding protein to a low albumin driven by inflammation addresses nothing.

The productive question is why it is low, and that means checking hs-CRP alongside, checking liver and kidney function, and looking at whether protein is being lost in urine.

In the other direction, I would not want low albumin dismissed as merely inflammatory either. The prognostic signal is real and it is a reason for a doctor to look further rather than to shrug.

One practical point: because albumin carries many medications, a genuinely low value changes how some drugs behave, which is a reason to mention it to a prescriber rather than a curiosity.

And a low albumin in someone with swelling in the legs or abdomen needs a doctor rather than a supplement.

So. What to actually do.

Tonight, at no cost. Find your albumin on your last comprehensive panel. It is almost certainly there and almost certainly unread.

Read it alongside an inflammation marker. hs-CRP is what tells you whether a low albumin is inflammatory rather than nutritional, and that distinction changes everything about what to do next.

Do not treat a low albumin by adding protein. If inflammation is driving it, protein addresses nothing, and that is the most common situation.

Ask about urine protein if it is low. Kidney loss is a real cause and it is checked with a simple urine test.

Take a persistently low value to a doctor. The prognostic signal is genuine and it warrants looking for the reason rather than watching.

Mention it to your prescriber if you take medication. Albumin carries many drugs, so a low value changes how some of them behave.

Turn up normally hydrated. Albumin is a concentration, so dehydration raises it and overhydration lowers it independent of anything real.

You are not short of protein because a protein came back low. You are a liver that switched production toward fighting something, and albumin was what it deprioritised.

Your body is not broken. It is blocked. And sometimes the block is an inflammatory process nobody named, showing up as a number everyone read as a diet problem.

Go read it next to your CRP.