Injected, oral or nasal: does the route matter for peptides?

For a peptide, the route decides how much of the dose actually reaches your blood, and the spread is enormous: injected semaglutide delivers 89% of the dose [2], an engineered oral tablet about 0.4–1% [3], and a plain desmopressin tablet roughly 0.16% of the same dose given intravenously [6]. That one figure, bioavailability, is why peptides are overwhelmingly injected, why oral semaglutide needed a purpose-built absorption enhancer to work at all, and why an oral BPC-157 capsule is a different proposition from an injected one.

The number to understand is bioavailability: the share of a dose that reaches the bloodstream intact. For an intravenous injection it is 100% by definition. For everything else, the question is how much gets lost on the way.

Why the gut is hostile to peptides

A peptide is a short chain of amino acids, and your digestive system exists to cut chains of amino acids into pieces. Stomach acid, pepsin and a battery of intestinal enzymes all attack peptide bonds. Anything that survives then meets the gut lining, a barrier of tightly joined cells coated in mucus that lets small, fat-soluble molecules through far more easily than large, water-loving ones such as peptides [1].

This is why oral versions of insulin, calcitonin, parathyroid hormone and vasopressin have been pursued for decades with limited success [1].

The official labels make the gap concrete. Desmopressin, a small synthetic peptide, is one of the few sold as a tablet. Its US label puts oral bioavailability at about 0.16% of an intravenous dose, and about 5% of what the same drug achieves through the nose [6].

Injection under the skin

Subcutaneous injection skips the gut entirely. The peptide sits in the fatty layer under the skin and seeps into small blood and lymph vessels. For injected semaglutide, the label gives an absolute bioavailability of 89% [2].

That is the reason injection is the default, not a preference. It is also why claims that an oral or topical version of a peptide works “just as well” deserve the same question every time: where is the data on how much reaches the blood?

The exception: oral semaglutide

Oral semaglutide was the first GLP-1 drug approved as a tablet, a milestone in a field where oral peptides had mostly stalled [1]. How it gets through shows how hard the problem is.

The tablet pairs semaglutide with an absorption enhancer called SNAC. Studies in dogs and people found that absorption happens in the stomach, in a small zone right next to the dissolving tablet. SNAC buffers the local acid to protect the peptide from enzymes and briefly helps it cross the stomach lining, through the cells rather than between them. Without SNAC, the absorption didn’t happen [4].

Even with all that, the label puts absolute bioavailability at roughly 0.4–1% for the original tablet strengths [3]. The label is strict about how it is taken: on an empty stomach, with no more than about 120 ml of water, and nothing else to eat, drink or swallow for at least 30 minutes [3]. Less water led to more absorption in the manufacturer’s own studies [3]. Semaglutide was also a good candidate to begin with: its elimination half-life is about a week [2], so a small daily fraction can build up.

It works. In PIONEER 1, a 26-week double-blind trial in 703 adults with type 2 diabetes, the 14 mg daily tablet lowered HbA1c by 1.1 percentage points more than placebo, and body weight by 2.3 kg more [5].

The lesson is that an oral peptide works once the molecule, the tablet and the dosing conditions are engineered around the gut’s defences. Semaglutide is the proof that the engineering can be done; the SNAC enhancer, the strict fasting conditions and the week-long half-life are what carry it.

Nasal sprays

The nose is a middle path. Its lining has a rich blood supply, and a drug absorbed there reaches the circulation without first passing through the liver. But how much gets in depends on the molecule’s size and chemistry, and the nose’s mucus-and-cilia system constantly sweeps material away. Poor absorption through the nasal lining is the main thing that stops more drugs using this route [13].

Again the labels give real numbers. Calcitonin salmon nasal spray reaches 3–5% of the bioavailability of an intramuscular injection [7]. Desmopressin does roughly 20 times better through the nose than as a tablet [6]. Nasal delivery beats swallowing, then, but still loses most of the dose compared with injection.

Some peptides marketed for brain effects are sold as nasal sprays or drops, on the theory that part of the dose reaches the brain directly. How much actually does, for any given peptide, is a separate question that needs its own measurements.

BPC-157: the oral claims versus the evidence

BPC-157 is the peptide most often sold as a capsule, and it has a specific argument behind that. It is described as a fragment of a protein found in gastric juice [10,11], and its original research group calls it a “stable gastric pentadecapeptide”, reporting unusual stability in gastric juice and effects when given by mouth or directly into the stomach in rats [9]. A 2026 review agrees it shows activity by oral, injected and topical routes in animals [10].

Here is what that does and doesn’t establish:

• Stability in gastric juice isn’t absorption. Surviving the stomach is step one. Crossing the gut wall into the blood is a separate hurdle, and that human exposure number is the one still missing for oral BPC-157.
• Local effects may not need absorption. If BPC-157 acts on the gut lining itself, an oral dose could matter for gut problems without much entering the bloodstream. That cuts against claims about tendons or joints elsewhere in the body.
• The formal pharmacokinetic work used injections. The one full study of how BPC-157 is absorbed, distributed and cleared, in rats and dogs, used intravenous and intramuscular dosing. Intramuscular bioavailability was about 14–19% in rats and 45–51% in dogs, and the peptide’s half-life in plasma was under 30 minutes [8].
• Human data so far come from small pilots. Reviews count three small human pilot studies, none controlled [10,11]. The only intravenous study in people was a safety pilot in two adults [12], and a 2026 review notes that the short half-life has been observed in people in that kind of preliminary two-person work [10].

A 2026 review sums up the gap: BPC-157’s human pharmacokinetics remain “critically undercharacterized”, and no pharmaceutical-grade formulation has been developed or validated for any route [10].

A quick comparison

What is still open

• How much of an oral BPC-157 capsule reaches human blood. This is the number that would settle the question, and it has not yet been published.
• Whether the gut-only effects seen in animal studies of oral BPC-157 carry over to people.
• For nasal peptides marketed for the brain, how much of a dose reaches brain tissue in humans.
• Whether the absorption tricks that carry semaglutide would work for other peptides. The 2018 study found SNAC’s effect was specific to the compound [4].

What this means

HPR’s position is that the route is not a delivery detail bolted onto a peptide claim; for a peptide it is the claim. Two products can name the same compound and the same milligram dose and still put wildly different amounts into your blood, and the labels show the range: 89% for an injected GLP-1, under 1% for the same drug engineered as a tablet, around 0.16% for a plain desmopressin tablet. A buying decision made on the ingredient rather than the route is a decision made on the wrong variable.

That gives a test we apply to any peptide sold by mouth. The seller either has exposure data in humans for that route, or has engineered the delivery so that a route with poor absorption is no longer the one being relied on. Semaglutide tablets clear both bars. An oral BPC-157 capsule sold on the strength of stability in gastric juice clears neither: stability is a property of the molecule in the stomach, not a measurement of what arrives in the blood. A reader can reasonably disagree with how much weight to put on animal data here, but the question to ask is fixed, and it is about exposure.

Our position on the nose is narrower. Nasal delivery sits between swallowing and injecting, it can beat a tablet, and it does not pass through the liver first. That is a genuine advantage. But a nasal spray marketed for effects in the brain still has to show the dose reaching the brain, and that is a measurement we want to see before treating the route as evidence of anything.

The practical rule follows. Ask for the pharmacokinetics before the anecdote, and expect a percentage. Apply the same discipline one step earlier to whatever you are buying: a capsule, spray or vial should come with batch testing for purity, sterility, endotoxin and heavy metals, and that testing is the bar a seller has to meet before any question about the route matters. Where a number exists, as it does for semaglutide and desmopressin, peptides can be discussed with real precision. Where it does not, the honest summary is a route claim waiting on its measurement, not a proven effect.

Educational content only — not medical advice. Many peptides discussed on HPR are not approved for human use. Talk to a qualified clinician before making any decision about your health.