Reconstitution math for GLP-1 self-titration: why dose calculators understate the risk that actually matters

What the arithmetic actually is, in two layers

Reconstitution math has two distinct layers. The first layer turns a vial of lyophilized peptide and a measured volume of bacteriostatic water into a concentration. If a vial holds 5 mg of semaglutide and is reconstituted with 2 ml of bacteriostatic water, the resulting concentration is 2.5 mg per ml, or equivalently 2500 mcg per ml. That step is multiplication and division, and a calculator handles it cleanly.

The second layer is harder. It converts a planned weekly dose in milligrams into a draw volume in milliliters, then converts that volume into syringe units on whatever syringe is at hand, then resolves how the syringe physically rounds the draw. The final dose the patient receives is the product of all three steps. The labeled GLP-1 products are designed by the manufacturer to skip most of this chain by shipping a calibrated pen . Compounded preparations push that chain back onto the user.

U-100 vs. U-50 syringes, and why the marking system is a trap

Insulin syringes are calibrated in units, not milliliters. A U-100 syringe is built so that 100 units equal 1 ml. A U-50 syringe is built so that 50 units equal 0.5 ml. The unit markings are useful only for U-100 insulin (100 IU per ml) or for arithmetic that explicitly does the conversion.

GLP-1 peptides are not insulin. They are not measured in international units. They are measured in milligrams or, derivatively, milliliters. When a user draws semaglutide from a 2.5 mg per ml vial up to the '20' mark on a U-100 syringe, they have drawn 0.20 ml, which contains 0.5 mg, not 20 mg and not 20 mcg. That conversion is correct, but it depends entirely on the user knowing that the U-100 marking represents a volume divided by 100, not a count of milligrams.

• U-100 syringe: 1 marking equals 0.01 ml. Common syringe sizes are 0.3, 0.5, and 1 ml, holding 30, 50, and 100 unit markings respectively.
• U-50 syringe: 1 marking equals 0.01 ml on the 0.5 ml size, but the syringe is labeled 0 to 50 instead of 0 to 100. The fluid volume per marking is the same; only the label differs.
• BD or similar U-40 syringes exist for veterinary insulin and should never be used for human GLP-1 work; the unit marking represents a different fluid volume per increment.

The FDA dosing-error alert specifically calls out the units-vs-milligrams misread as a recurring failure mode in compounded GLP-1 use . This is the single most common error class.

Syringe rounding, and the under-discussed compounding effect

A U-100 syringe with 100 markings on a 1 ml barrel resolves to 0.01 ml per marking. In practice, the eye reads to about half a marking, or 0.005 ml. At a 2.5 mg per ml concentration, that resolution corresponds to about 12.5 mcg, which is small relative to a typical GLP-1 weekly dose. At a higher concentration (such as a 5 mg per ml vial reconstituted from the same 5 mg powder using only 1 ml of bacteriostatic water), the same half-marking corresponds to 25 mcg of error. The rounding error scales linearly with the concentration.

The compounding effect comes from titration ladders. STEP-1 used a 16-week titration from 0.25 mg to 2.4 mg weekly, with the dose roughly doubling every four weeks during ramp . SURMOUNT-1 used a similar five-step ladder for tirzepatide . A user who self-titrates from a single high-concentration vial without dilution carries a dosing-error variance that grows with every ladder step, because the syringe-marking resolution is the same in absolute volume terms but a smaller fraction of the larger draw.

The mitigation is concentration-aware: lower concentrations make rounding error smaller relative to the dose, but they also push the draw volume past the 0.5 ml or 1 ml syringe limit at higher dose tiers. A practical reconstitution decision picks a concentration that keeps every step of the planned titration within a draw volume between 0.10 ml and 0.40 ml, where rounding error sits between 1 and 5 percent of the dose. Outside that band, error magnitudes climb.

Bacteriostatic water volume, and what the label actually specifies

Bacteriostatic water for injection is sterile water containing 0.9 percent benzyl alcohol as a preservative. It is appropriate for multi-dose vials drawn over multiple administrations because the preservative limits microbial growth between draws . Sterile water for injection has no preservative and is intended for single-use reconstitution.

The volume of bacteriostatic water added to a lyophilized vial determines the final concentration. A 5 mg vial reconstituted with 1 ml of bacteriostatic water yields 5 mg per ml. The same vial reconstituted with 2 ml yields 2.5 mg per ml. Both concentrations deliver the same total drug per vial; the choice trades draw-volume convenience against syringe-rounding precision. Approved GLP-1 products do not give the user this choice because the manufacturer ships a fixed-concentration cartridge.

What a defensible titration plan looks like, drawing from published trial design

STEP-1 used a fixed-step titration with 4 weeks at each dose level: 0.25, 0.5, 1.0, 1.7, and 2.4 mg weekly. SURMOUNT-1 used a similar 4-week dose-step pattern for tirzepatide at 2.5, 5, 7.5, 10, 12.5, and 15 mg . The 4-week step is not arbitrary; it gives time for tachyphylaxis at each receptor occupancy level, blunts nausea, and lets clinical assessment catch dose-related adverse effects before the next escalation.

A self-administration plan that compresses these intervals to two weeks doubles the total integrated escalation rate without doubling the corresponding tolerance. The trial GI-tolerability data (nausea, vomiting, constipation, diarrhea) tracks the rate of escalation as much as the absolute dose level. Faster ramps have higher acute discontinuation rates in published trial subgroup analyses.

• Plan the full ladder before the first injection. Reconstitution concentration should keep every planned dose between 0.10 ml and 0.40 ml of draw volume.
• Document the labeled-product equivalents at each ladder step for cross-reference, including the FDA-approved monthly cadence for any approved equivalent.
• Track total weeks-on-dose at each tier rather than calendar weeks; missed weeks count.
• Plan a stop or step-down threshold for nausea, vomiting, or any GI event severe enough to interfere with hydration. The labeled products specify this; self-administration plans frequently do not.

Where the dose calculator gets the answer right but misses the risk

• Vial-to-vial variance: compounded vials may not deliver the labeled mg per vial within tight tolerances. The calculator assumes the label is correct.
• Salt-form variability: some compounded preparations have used different salt forms of the active ingredient, with different molecular weights and therefore different mg-equivalent doses for the same volume.
• Storage decay: peptide concentration drops over weeks of refrigerated storage; the calculator assumes the day-zero concentration applies forever.
• Air-gap dead volume: a 1 ml syringe holds about 0.05 ml of dead volume in the hub and needle. Drawing 0.10 ml means 0.10 ml of the contents reaches the syringe barrel, but only 0.10 ml minus dead-volume reaches the patient. With low draws, this is a meaningful percentage of the dose.
• User-perception variance: the eye reads 0.005 ml differently in different lighting and at different barrel angles. Trial patients use pens for a reason.