Kidneys

What filtration actually is

Filtration here is a physical process rather than a chemical one. There is no recognition step, nothing that inspects a molecule and decides.

Blood arrives at a tight knot of capillaries called the glomerulus and is held there at a higher pressure than anywhere else in the circulation, because the vessel leaving the knot is narrower than the one arriving. That pressure pushes fluid out through a three-layered wall, and the wall behaves as a sieve. Small things go through, large things do not. Water, salts, glucose, urea and creatinine pass easily. Blood cells and the large proteins do not.

This is why protein in the urine means something structural is wrong. Albumin sits just the wrong side of the cut-off, a little too large to pass in any quantity through an intact barrier. When albumin starts appearing in urine, it is not that the kidney has decided to let it go. It is that the sieve itself has been damaged.

One part of the standard account is less settled than is usually taught. Textbooks add that the barrier also carries a negative charge and repels negatively charged molecules like albumin, holding back more of it than size alone would predict. Some experiments support that and others do not, and animals engineered to lose much of the barrier's negative charge turn out not to leak protein. Selection by size and by shape is secure. Selection by charge is still argued about.

The other thing to hold in mind before the numbers below is that the sieve is deliberately crude. It lets through an enormous volume containing a great deal you cannot afford to lose. The precision happens afterwards, further down the tubing, where almost all of it is taken back.

Filtering waste is only part of the job. The kidneys also decide how much water and salt you keep, hold your blood acidity in a narrow band, help set your blood pressure, activate vitamin D into its usable form, and release the hormone (Erythropoietin) that tells your bone marrow to make red blood cells. A kidney that is struggling is rarely a problem confined to the kidney.

The kidney as a gland

That part of the job is usually given as a list, which hides how good the design is. Three of those tasks are hormonal, and each one exists because the kidney happens to be the right place to notice something.

Erythropoietin (EPO), the red cell signal. Scattered in the tissue between the tubules, in the outer part of the kidney, is a population of cells that do little except sense oxygen. The sensing works by subtraction. A set of enzymes continuously tags a particular signalling protein for destruction, and those enzymes need oxygen to do it. When oxygen falls the tagging slows, the signalling protein survives instead of being destroyed, and it switches on the gene for EPO. More EPO, more red cells, more oxygen carried, and the sensing quietens down again.

The kidney is a sensible place for that sensor, because oxygen delivery depends both on how much blood is arriving and on how much oxygen the blood is carrying, and an organ taking a fifth of the circulation is exposed to both at once.

Renin, the pressure signal. Specialised cells wrapped around the small artery feeding each filtering unit sense the pressure inside it. If pressure falls they release renin, which begins a cascade that narrows blood vessels and instructs the body to hold on to salt and water. The kidney's answer to low pressure is to raise pressure.

That is exactly right after blood loss. It is less helpful when the low pressure the kidney is sensing comes from its own narrowed artery, because the whole body then runs at a higher pressure to satisfy one organ. That particular situation accounts for only a small share of high blood pressure, somewhere between one and five in every hundred cases, but it is the clearest illustration of something much larger.

Why the kidney sits at the centre of blood pressure

Over months and years, the kidney has the final say. In the short term blood pressure is set by how hard the heart pumps and how tight the vessels are. Over the long run it settles wherever the balance lands between the salt and water you take in and the salt and water the kidney lets go. Everything else pushes the number around temporarily, and on that account the kidney decides where it comes to rest. That is the dominant view rather than a settled one. Changes starting in the blood vessels or in the nerves supplying them can shift pressure over the long run without the kidney handling salt any differently.

That is why several of the main classes of blood pressure medication work on this system rather than on the heart or the vessels. Diuretics make the kidney release more salt and water. ACE inhibitors, angiotensin receptor blockers and aldosterone antagonists all interrupt the renin cascade at different points.

Most high blood pressure has no single identifiable cause and is called primary, or essential, hypertension. That name is often read as nobody knowing what is happening, which is not quite it. The balance point has shifted and the kidney is holding pressure higher than it needs to. What is missing is one thing to blame.

Vitamin D activation. When consumed through diet or produced through the skin from UVB light, Vitamin D is not active. The liver performs one modification and the kidney performs the second. It is the second one that produces the active hormone. The Nutrients section further down covers it, and it is part of why kidney trouble and calcium trouble tend to turn up together.

What the profile looks at

A substance is useful for measuring filtration if it meets four conditions. The body produces it at a steady rate, the sieve passes it freely, the tubing downstream does not take it back, and nothing adds to it on the way out. Creatinine meets three of the four.

It is a waste product of normal muscle turnover, produced at a rate set mainly by how much muscle you carry, filtered freely, and not reabsorbed. The condition it fails is the last one because the tubule actively secretes a further amount on top of what was filtered. This amount also grows as filtration falls. So the worse the filtration gets, the more creatinine the kidneys push out, making it look better than it actually is. It also means anything competing for that same transporter raises creatinine while filtration is entirely unchanged, which matters before you read a single number as a decline.

Creatinine's dependence on muscle is the larger limitation, and the Muscle section further down this page covers it properly, along with what it means for anyone heavily built, anyone taking creatine, and anyone who has recently eaten a lot of red meat.

Urea is how the body disposes of nitrogen. Protein is broken down, ammonia is produced, and the liver converts it into urea because urea is safe to carry in the blood. The kidneys then clear it. That path leaves urea sitting at the junction of 4 different things: protein intake, liver function, hydration and filtration. It moves for any of them. It is read next to creatinine rather than instead of it precisely because the 2 have different ways of being wrong.

eGFR is the one most people look at first, and it is not measured. It is estimated glomerular filtration rate, an equation run on your creatinine, your age and your sex. Everything above about creatinine therefore applies to it too. It inherits every one of creatinine's problems and adds the assumptions of the equation on top. Which equation, and how much a result has to move before it means anything, are in the Biomarker Guide.

Cystatin C is a small protein produced at a near-constant rate by essentially every cell with a nucleus, so it does not depend on muscle at all. It is filtered freely and then broken down by the tubule rather than returned, so almost none comes back into the blood. It is useful in exactly the situations where creatinine is least reliable, which is why anyone carrying a lot of muscle should probably test it. It can be added to any venous test.

International kidney guidance updated in 2024 now recommends that where cystatin C is available, filtration should be estimated from creatinine and cystatin C together rather than from creatinine alone. That is a strong recommendation, and it sits some way ahead of what most people are actually given.

Uric acid, also an add-on, is a third waste product the kidneys clear, this one from the breakdown of purines, the building blocks of DNA, which come both from your own cell turnover and from food. It is unusual among waste products in that it causes trouble by coming out of solution rather than by building up quietly. Above a certain concentration it crystallises, in joints as gout and in the urinary tract as one type of kidney stone. It also tracks closely with insulin resistance and cardiovascular risk, though whether it drives the risk or just travels alongside it is not yet known for certain. What has changed is the emphasis. Urate is now seen as far more of a metabolic marker than a dietary one, and diet is a smaller lever on it than the standard advice about red meat implies.

What a blood profile does not cover

Kidney damage is staged on two main axes rather than one, and this is where the numbers are worth publishing because they are national guidance rather than anybody's opinion.

The first axis is filtration. Reduced filtration is defined at an eGFR below 60, and that figure comes from the National Institute for Health and Care Excellence (NICE) and from international kidney guidance, which use the same staging.

The second axis is how much albumin is leaking into the urine, measured as a urine albumin-to-creatinine (ACR) ratio. A confirmed ACR of 3 mg/mmol or more is abnormal, and it is enough to define kidney disease on its own, even with a completely normal eGFR. Guidance is explicit that risk should be read from the two axes together rather than from filtration alone.

That second one is a urine test, and we do not sell it.

Home urine dipsticks are widely available, so here is what they can and cannot do. An ordinary strip reads total protein rather than albumin specifically, and it is not sensitive at the concentrations that matter. The range that defines early kidney disease will usually read negative on one. NICE says not to use reagent strips to identify protein in the urine unless they can specifically measure albumin at low concentrations and report the result as a ratio.

So an ordinary clear dipstick is not reassurance.

What you can actually do about it. Strips that meet that standard do exist. They are sold as albumin-to-creatinine or microalbumin strips rather than plain protein strips, and some are designed to be photographed with a phone app that returns an actual ratio. Read the packaging and check it says albumin and gives a ratio, because the cheap protein-only strips are far more common and will not do this job.

Otherwise a lab ACR is a standard test any GP can arrange, and it is also sold privately by most testing companies. It is one of the cheapest tests there is, and it covers the half of the picture a blood panel cannot.

It matters because albumin in the urine can appear years before the filtration rate moves. In kidney disease caused by diabetes it is very often the first sign there is. A normal eGFR on its own does not rule kidney disease out.

Anyone with diabetes, high blood pressure, cardiovascular disease, gout, a previous episode of acute kidney injury, or a family history of kidney disease should have a urine albumin-to-creatinine ratio alongside a blood test, and it is a standard test a doctor can arrange. A blood profile covers one of the two axes. It does not cover both.

What changes when the kidneys are under strain

Kidney function declines quietly. They have a considerable inbuilt spare capacity, so a substantial amount can be lost before anything is noticeable. When symptoms do appear, they tend to be general rather than specific: fatigue, swelling around the ankles and feet, changes in how often or how much you pass urine, poor sleep, itchy skin, a loss of appetite. Every one of those has many other explanations.

That silence is the argument for measuring rather than waiting to feel something.

One of those symptoms is worth tracking, because it links this section to the one on blood cells. Fatigue in reduced kidney function is often anaemia, and two mechanisms run at once. The kidney makes less EPO, so the marrow is told to produce less. And hepcidin, the iron gatekeeper described in the Iron section, is itself cleared by the kidney, so as filtration falls it accumulates and iron is held in storage (as ferritin) instead of being released for red cell production. Less signal and less available iron, from one cause.