Thigh, Abdomen or Arm: Does GLP-1 Injection Site Change Absorption?

Abdomen, thigh, upper arm. Every GLP-1 label lists the same three subcutaneous sites, and patients reasonably assume the choice must matter — that the thigh absorbs slower, or that the belly “works better.” For older, short-acting insulins that intuition was well founded. For the once-weekly GLP-1 receptor agonists it largely is not. The question was tested directly in a randomized crossover study of an injectable GLP-1 receptor agonist, and the sites were close enough that the authors concluded no dose adjustment was needed when switching between them^[1].

What was actually measured

The cleanest evidence comes from a randomized, open-label crossover study that gave the same participants the same dose in the abdomen, the thigh and the upper arm on separate occasions, then compared the resulting drug exposure^[1]. That design is the right one for this question: each person acts as their own control, so differences in body composition between people cannot masquerade as differences between sites. The conclusion was that bioavailability across the three sites was comparable, and that patients could rotate between them without changing the dose.

The same picture holds for the long-acting agents through a different route of evidence. Population pharmacokinetic analyses of dulaglutide pooled data across its clinical trial program and did not identify injection site as a factor requiring dose adjustment^[2]. A systematic review of semaglutide's clinical pharmacokinetics describes absorption slow enough, and a half-life long enough, that the drug is dosed once weekly^[3] — and tirzepatide's population pharmacokinetic model supports the same weekly interval^[4].

What this does and does not license

• Choosing the thigh because it is easier to reach is reasonable. The evidence does not support the idea that you are getting less drug there. Comfort, reach and visibility are legitimate grounds to prefer a site.
• Rotating sites is still sound practice. The reason is tissue, not pharmacokinetics — repeated injection into the same square inch is what produces local lumps, bruising and irritation. Rotation protects the skin; it is not correcting an absorption problem.
• Subcutaneous means subcutaneous everywhere. The finding is about which fatty area you choose, not about depth. An injection that goes intramuscular because the site was thin or the needle long is a different event from one delivered correctly into fat, and that is a technique question rather than a site question.
• It is not a license to improvise sites. The three named areas are the ones studied and labeled. “The evidence says site does not matter much” applies among those three, not to anywhere you can reach.
• Do not read it as license to change your own dose. The crossover result says the same dose behaves similarly across sites — which is an argument for keeping your dose constant, not for adjusting it when you switch.

The thigh specifically

The thigh attracts the most questions, probably because it is the site people move to when abdominal injections start to feel repetitive or uncomfortable. The practical points are unremarkable: use the front and outer portion of the upper leg where there is a fold of fat to pinch, stay away from the knee and the inner thigh, and rotate the exact spot within that area rather than returning to the same point weekly. Nothing in the pharmacokinetic literature suggests you should expect a different result from the drug there^[1].

One caveat is worth stating honestly, because it is a genuine limit of the evidence rather than a reassurance: these studies enrolled adults across a range of body types, but they were not designed to answer what happens at the extremes of very low or very high subcutaneous fat at a specific site. If you have little fat at the thigh, the relevant risk is depth — an unintentionally intramuscular injection — and that is worth raising with a clinician rather than resolving from a table.