Liraglutide is the odd compound in this section: for two of these three conditions it is not off-label interest at all, it is the approved indication with two decades of trial data behind it. The third — PCOS — is off-label, and is also where this molecule has more randomised evidence than any other GLP-1 agonist.
Theorized — the randomised evidence belongs elsewhere in the class
Smoking cessation — class interest, no trial for this agent
Pathophysiology
Nicotine dependence is maintained by mesolimbic dopamine signalling. Quitting also produces an average weight gain of several kilograms, and fear of that gain is a measured barrier to attempting cessation and a measured cause of relapse.
Mechanistic rationale
The randomised evidence in this class is two small trials. Dulaglutide alongside varenicline did not improve abstinence — 63% against 65% — but prevented post-cessation weight gain, −4.6 kg against +0.5 kg. Exenatide alongside a nicotine patch did improve abstinence, 46.3% against 26.8%. Semaglutide has a large observational signal whose own investigators said it does not justify off-label use. Nothing has been tested for this agent, and in every trial the GLP-1 was added to a treatment that already works rather than replacing it.
Community reports
The consistent finding across the randomised literature is the weight, not the quitting. Post-cessation weight gain is real, specific and worth addressing; improved abstinence is not established, and the whole randomised evidence base is roughly 410 participants.
Components carrying the argument: GLP-1 receptor — class inference only
Approved — randomised outcome trials
Type 2 diabetes — the approved indication
Pathophysiology
Type 2 diabetes is a progressive disorder of insulin resistance plus beta-cell failure. Glucose control alone does not reliably reduce cardiovascular death; the reason this class changed practice is that it does.
Mechanistic rationale
Liraglutide is glucose-dependent in its insulinotropic action, which is why it lowers HbA1c without the hypoglycaemia risk of sulphonylureas or insulin — unless it is combined with them. The LEAD programme established the glycaemic effect; LEADER (9,340 participants with type 2 diabetes at high cardiovascular risk) established a reduction in major adverse cardiovascular events and in cardiovascular death versus placebo. That result, not the HbA1c number, is why it sits high in the treatment algorithms.
Community reports
The daily schedule is the recurring complaint and the recurring advantage. People who could not tolerate weekly semaglutide sometimes report that liraglutide is manageable, because a missed or reduced daily dose self-corrects within a day rather than sitting in the system for a week.
Components carrying the argument: GLP-1 receptor — glucose-dependent insulin secretion, glucagon suppression, gastric emptying delay
Randomised human trials — off-label
PCOS — the best-studied agent in the class
Pathophysiology
Insulin resistance drives hyperinsulinaemia; hyperinsulinaemia drives ovarian androgen production and suppresses SHBG, which raises free testosterone. Weight loss and improved insulin sensitivity unwind that loop from both ends.
Mechanistic rationale
More randomised PCOS data exists for liraglutide than for any other agent in this class, largely because it has been available longest. Trials — mostly small, mostly 12 to 32 weeks — have reported weight and waist reduction, improved insulin sensitivity, and in several, restored menstrual regularity, with combination against metformin studied more than either alone. The evidence is real; it is also thin next to the diabetes and obesity literature, and none of it is powered for pregnancy or live-birth outcomes.
Community reports
Community reporting concentrates on cycle regularity returning within a few months and on the emotional weight of that after years of irregularity. Reporting is self-selected — people for whom nothing changed rarely post.
Components carrying the argument: Weight loss plus a direct improvement in insulin sensitivity
Actionable
Fertility returns — and pregnancy is a contraindication
Pathophysiology
Anovulation in PCOS is frequently reversible. When ovulation resumes, so does the possibility of conception — often without warning, in people who had stopped using contraception because they believed they could not conceive.
Mechanistic rationale
This class is contraindicated in pregnancy and the labelling asks for discontinuation in advance of a planned pregnancy. Liraglutide clears faster than the weekly agents, which shortens the washout, but the point stands: the treatment that restores fertility is the treatment that must stop before that fertility is used.
Community reports
Unplanned conception on incretins is a recurring theme in PCOS communities, frequently framed as a happy surprise. The clinical framing is less relaxed.
Components carrying the argument: A predictable consequence, not a side effect
Approved — randomised trials
Obesity — approved at a higher dose, as a different product
Pathophysiology
Obesity is defended physiologically: energy restriction lowers the metabolic rate and raises hunger signalling, which is why most weight loss reverses.
Mechanistic rationale
The obesity indication uses 3.0 mg daily; the diabetes indication tops out at 1.8 mg. Same molecule, different product, different titration schedule, different label. The SCALE programme supported the obesity approval. Newer weekly agents produce substantially more weight loss, which is why liraglutide has largely been displaced for this use — not because it stopped working.
Community reports
People who have moved from liraglutide to a weekly agent almost universally report more weight loss; a minority report they preferred the daily version for tolerability and went back.
Components carrying the argument: Appetite and satiety signalling in the hypothalamus
What actually has evidence for these conditions
Type 2 diabetes has an established treatment algorithm and it does not start here. Metformin remains first-line and is inexpensive; SGLT2 inhibitors have their own cardiovascular and renal outcome data; this class is added for people with established cardiovascular disease, chronic kidney disease, or when weight is central to the picture. Structured education, sleep, resistance training and dietary change are not the polite preamble to drug therapy — they change the trajectory of the disease and they compound with everything above.
For PCOS: metformin for the metabolic phenotype, combined hormonal contraceptives for hyperandrogenism and cycle control, letrozole first-line for ovulation induction, inositol with reasonable evidence and a benign safety profile. A 5–10% weight reduction alone restores ovulation in a meaningful proportion of people.