What a Low Dose Actually Does

The argument about lower doses has been conducted almost entirely without numbers. Here is one: adults with obesity and without diabetes, given tirzepatide at 2.5 mg for four weeks then 5 mg, lost 8.2 kg — 7.3% of body weight — over twelve weeks, with 46.1% reaching at least 5%.^[1]

What the low dose produced

• Weight: −8.2 ± 4.9 kg, a 7.3% reduction, with BMI down 2.8 kg/m².
• Responders: 46.1% reached the 5% threshold conventionally used as clinically meaningful.
• HbA1c: 5.6% to 5.4%. LDL cholesterol: 113 to 106 mg/dL. Triglycerides: 123.6 to 119.2 mg/dL. HDL and kidney function unchanged.
• Tolerability: nausea in 7.8%, discontinuation in 10.4%.^[1]

That is a real result from the bottom two rungs of a five-rung ladder, and it is the figure missing from most conversations about whether people need to climb all the way.

What it does not settle

Twelve weeks is the steep part of any weight curve. Loss is fastest early and slows, so multiplying this out to a year would overstate the outcome badly — the full-dose trials ran 72 weeks precisely because the shape matters.

There is also no control group and no comparison arm at a higher dose. This tells you what happened at 5 mg. It does not tell you what those same people would have achieved at 15 mg, which is the actual question anyone weighing a lower dose is asking.

A number for the low dose is not a comparison between doses. Only a randomized trial gives you that.

And 5 mg is a labeled dose. It is not microdosing, which means dosing below the labeled range — a distinction we keep drawing because the marketing keeps collapsing it. See what evidence exists for microdosing and the randomized dose-reduction trial.

The detail worth taking away

Discontinuation was 10.4%, and it occurred mainly among people previously treated with GLP-1 receptor agonists.^[1]

That is the second study in short order pointing the same way. The SEMALEAN cohort found attenuated responses in people with prior GLP-1 exposure. Here, prior exposure predicted stopping.