Semaglutide

Semaglutide is a modified version of the gut hormone GLP-1, engineered to last a week in the body instead of minutes. It was approved for type 2 diabetes first and for obesity second, and the obesity trials are what changed the field.

The weight trials. STEP 1 randomised 1961 adults with overweight or obesity and no diabetes. At 68 weeks, mean weight change was -14.9% on semaglutide 2.4 mg against -2.4% on placebo, a difference of 12.4 percentage points; 86.4% lost at least 5%, 69.1% at least 10%, and 50.5% at least 15% [1]. STEP 5 extended that to two years and found it held: -15.2% versus -2.6% at week 104 [4]. STEP 8 compared it head to head with daily liraglutide 3.0 mg and won decisively, -15.8% versus -6.4% [5]. In adolescents with obesity the effect was of the same size [6]. STEP 3 added intensive behavioural therapy on top [7], and STEP 4 showed that continuing the drug after an initial 20-week run-in keeps weight falling while switching to placebo reverses it [8]. STEP UP tested 7.2 mg, above the licensed dose, and got -18.7% against -15.6% for 2.4 mg [9].

The cardiovascular trial. SELECT enrolled 17,604 people with overweight or obesity and established cardiovascular disease but no diabetes. Major adverse cardiovascular events occurred in 6.5% on semaglutide against 8.0% on placebo, a hazard ratio of 0.80 [2]. This is the result that made semaglutide a cardiovascular drug rather than a cosmetic one, and it is the reason obesity treatment is now reimbursed in places it was not. Kidney outcomes moved in the same direction [10][11], and the long-term weight effect in that population held over years [12]. In type 2 diabetes, SUSTAIN-6 had already shown a hazard ratio of 0.74 [13].

Beyond weight and heart. In obesity with knee osteoarthritis, 68 weeks produced -13.7% weight and a large improvement in WOMAC pain score [14]. In obesity-related heart failure with preserved ejection fraction, pooled STEP-HFpEF data showed a 7.5-point gain in the KCCQ symptom score [15].

The tablet. Oral semaglutide at 50 mg daily gave -15.1% at 68 weeks in OASIS 1, essentially matching the injection [16]; a 25 mg dose gave -13.6% at 64 weeks [17]. The catch is the dosing ritual: empty stomach, small sip of water, nothing else for half an hour.

What happens when you stop. The STEP 1 extension followed 327 participants for a year after withdrawal. They had lost 17.3% on drug and regained 11.6 percentage points of it, ending at a net 5.6% below baseline [3]. Cardiometabolic improvements reverted along with the weight. This is the single most important practical fact about the drug: it treats obesity for as long as it is taken.

GLP-1 is released from the gut after a meal and does several things at once: it stimulates glucose-dependent insulin secretion, suppresses glucagon, slows gastric emptying and signals satiety in the hypothalamus and brainstem. Native GLP-1 is destroyed by DPP-4 within minutes.

Semaglutide is human GLP-1 with an alanine-to-alpha-aminoisobutyric-acid substitution at position 8 that blocks DPP-4 cleavage, and a C18 fatty diacid attached through a spacer that binds albumin reversibly. Albumin binding is what buys the one-week half-life [18].

For weight, the dominant mechanism is central appetite suppression rather than the slowed stomach - although delayed gastric emptying is real, contributes to early satiety, and is the source of both the nausea and the anaesthetic problem [19]. The glucose-dependent nature of the insulin effect is why semaglutide alone rarely causes hypoglycaemia.

The cardiovascular benefit in SELECT is larger than weight loss alone easily explains, and the mechanism is still argued over: some combination of weight, blood pressure, inflammation and direct vascular effects [2][20].

Two routes with quite different pharmaceutics. The injection is a weekly subcutaneous dose from a pen, marketed as Wegovy for weight and Ozempic for diabetes; escalation runs from 0.25 mg to 2.4 mg over about 16 weeks.

The tablet is the interesting piece of engineering. Peptides are not absorbed from the gut, so oral semaglutide is co-formulated with sodium N-(8-[2-hydroxybenzoyl]amino) caprylate, which raises local gastric pH and transiently permeabilises the epithelium. Absorption is still poor and highly variable, which is why the daily oral dose is 25-50 mg against 2.4 mg injected weekly, and why the empty-stomach rule is not optional [16][17].

Gastrointestinal effects are the defining tolerability problem: nausea, vomiting, diarrhoea and constipation, mostly during escalation. In STEP 1, gastrointestinal adverse events led 4.5% to discontinue against 0.8% on placebo [1]; at the higher 7.2 mg dose, nausea reached 70.8% [9]. In SELECT, 16.6% discontinued the drug for adverse events against 8.2% on placebo [2].

Anaesthesia and sedation. Delayed gastric emptying means residual stomach contents after standard fasting. Observational work found more residual gastric content in GLP-1 users before anaesthesia [19]; a systematic review of pulmonary aspiration found the absolute risk small but the association present [21]. Tell an anaesthetist.

Lean mass. Large weight loss takes muscle with it, and GLP-1-based weight loss is no exception: fat-free mass typically accounts for around a quarter to 40% of the total, broadly in line with what diet-induced loss of the same magnitude produces [22][23][24]. Whether it is worse than equivalent dieting is genuinely contested; resistance training and adequate protein are the standard mitigations [22].

Eyes. SUSTAIN-6 found more diabetic retinopathy complications on semaglutide (HR 1.76), attributed at the time to rapid glucose lowering [13]. Separately, a possible association with non-arteritic anterior ischaemic optic neuropathy was raised in 2024. The follow-up evidence is mixed: a multicentre observational study and a multinational population study have reported differing magnitudes, and the question is not settled [25][26].

Tumours. GLP-1 agonists carry a rodent thyroid C-cell tumour warning. A systematic review of once-weekly agents found no clear human signal [27].

Pregnancy. Not for use in pregnancy, and the long half-life means stopping well in advance.

Delayed gastric emptying can slow the absorption of oral drugs taken at the same time, which matters most for narrow-therapeutic-index medicines [19].

With insulin or a sulfonylurea, the risk of hypoglycaemia rises and doses of those usually need reducing; semaglutide alone rarely causes it because its insulin effect is glucose-dependent [13].

Stacking with other GLP-1 or incretin agonists adds nothing except adverse effects. Combining with an amylin analogue is a different matter and is being developed deliberately: cagrilintide plus semaglutide gave -20.4% at 68 weeks in REDEFINE 1 [28]. Network meta-analyses place semaglutide behind tirzepatide and retatrutide and ahead of the older agonists [29][30].