Carnosinase and absorption

The short version

A person who swallows carnosine absorbs most of it intact through the gut. Serum carnosinase (CN1) then splits it into beta-alanine and histidine within minutes. Carnosine’s half-life in human serum is under five minutes (Nutrients 2023 review). Two older claims are both wrong: that carnosine is “not absorbed”, and that it is “well absorbed” in the sense of staying in the blood. It is absorbed well and cleared fast.

Where an oral dose goes

1. Gut wall. The small intestine takes carnosine up intact through the peptide transporter PEPT1. The gut lining also carries its own carnosinase, highest in the jejunum, which splits some of it on the way through (Sadikali 1975).
2. Blood. CN1 splits most of what arrives. Plasma carnosine is usually undetectable minutes after a dose. Older studies under-measured it because the enzyme kept working in the blood sample after collection (Gardner 1991). After a 200 g beef meal supplying about 250 mg carnosine, plasma carnosine was detectable at 15 minutes, peaked at 2.5 hours and was gone by 5.5 hours (Park 2005).
3. Tissues. Cells take up intact carnosine through the transporter PEPT2 and related carriers. In a rat kidney cell line, uptake was active and efflux slow (Jappar 2008). In mice without PEPT2, carnosine was cleared from blood faster and reabsorbed far less by the kidney (Kamal 2009). Mice lack serum carnosinase, so the balance between uptake and breakdown differs from people.
4. Brain. After a large single dose in five volunteers, brain carnosine on MR spectroscopy rose at one hour and returned to baseline by five (Ali 2025). No placebo group.
5. Kidney and urine. Carnosine is filtered freely. Gardner 1991 recovered up to 14 per cent of a 4 g dose intact in urine within five hours. The kidney tubule reabsorbs part of it through PEPT2.

So plasma carnosine falls for three reasons at once: breakdown by CN1, uptake into cells, and loss in urine. A low plasma level does not by itself show how much reached tissue. No human study has measured what share of an oral dose ends up as intact carnosine in tissue, as beta-alanine and histidine, or in urine.

The absorption studies

Gardner and colleagues gave carnosine by mouth to human volunteers in 1991 and measured what reached the urine and blood (Gardner 1991). Park and colleagues measured plasma carnosine after a beef meal (Park 2005). A dose-escalation study gave healthy volunteers up to 15 g a day and measured plasma and brain levels (Ali 2025). The Bellia 2014 review covers the carnosinase enzymes in detail (Bellia 2014).

After people eat chicken-derived dipeptides, only trace anserine appears in blood and carnosine is not detected (2026 preprint, not peer reviewed).

Why beta-alanine instead

Because of CN1, oral carnosine does not raise muscle carnosine in people. Beta-alanine does, because muscle makes carnosine from it and beta-alanine is the limiting ingredient (Australian Institute of Sport fact sheet). See Beta-alanine, the precursor.

What the clinical trials did anyway

Despite the enzyme, trials have given carnosine itself by mouth, at 0.5 to 2 g a day, in autism, schizophrenia, prediabetes and diabetes. How carnosine acts when most of it is broken down in minutes is an open question. Some authors suggest that the breakdown products, or brief pulses of intact carnosine taken up by tissue before breakdown, or an action in the gut, account for the results. Others suggest the trials’ results reflect their small size. The study pages in Humans state what each trial measured and found.

Variation between people

Serum CN1 activity varies widely between people. A repeat polymorphism in the CNDP1 gene, the amount of enzyme in serum and age all contribute (Separations 2023 review). People with low activity have more carnosine in their blood after a dose. See Breakdown by carnosinase.