GLP-1 drugs explained: a doctor-first guide
If you've heard about Ozempic, Wegovy, Mounjaro, or Zepbound and want to understand what they actually do — without the breathless headlines or the Instagram pseudoscience — this is a long read written by people who cite primary sources and refuse to tell you what to do with your body. We are not your doctor. We will say that often. You should also see one.
What are GLP-1 drugs, biologically?
Glucagon-like peptide-1 (GLP-1) is a hormone that your gut produces in small amounts whenever you eat. It does four useful things: it signals your pancreas to release insulin, it slows the rate at which your stomach empties into your small intestine, it dampens appetite signals in the hypothalamus, and it improves how your body handles blood sugar over the longer term. GLP-1 receptor agonists are synthetic molecules that mimic this natural hormone, but with a much longer half-life — so a once-weekly injection keeps the effect running continuously instead of in brief post-meal pulses.
The class includes:
- Semaglutide — sold as Ozempic (diabetes indication), Wegovy (weight indication) and Rybelsus (oral form). The most-studied molecule in the class.
- Tirzepatide — sold as Mounjaro and Zepbound. A "dual agonist" that activates both GLP-1 and GIP receptors. In head-to-head trials it has produced larger weight-loss numbers than semaglutide.
- Liraglutide — sold as Victoza and Saxenda. An older molecule, daily injection instead of weekly.
- Dulaglutide — sold as Trulicity. Mostly diabetes use.
- Retatrutide — in Phase 3 trials. A triple agonist (GLP-1 + GIP + glucagon). Early data suggests even bigger weight loss than tirzepatide.
What the evidence actually shows
The good — large, replicated, peer-reviewed
- Weight loss: 15–20% of total body weight over 12–18 months on average. The landmark trials are STEP-1 (semaglutide, NEJM 2021) and SURMOUNT-1 (tirzepatide, NEJM 2022). These are not small-sample wellness studies; they are multi-thousand-person randomised controlled trials.
- HbA1c reduction in type 2 diabetes: consistently 1–2% absolute reduction. This is large, and translates to real reductions in long-term complications.
- Cardiovascular events: the SELECT trial (semaglutide, Lincoff et al., NEJM 2023) showed a 20% reduction in major adverse cardiac events in obese non-diabetic adults with prior cardiovascular disease.
- Kidney protection: FLOW trial showed semaglutide slowed kidney disease progression in type 2 diabetes.
- Sleep apnoea: SURMOUNT-OSA showed tirzepatide reduced apnoea-hypopnoea events.
- Liver disease: improvement in MASH (formerly NASH) biomarkers in semaglutide trials.
These benefits are real and consistent. We don't downplay them, and neither should anyone else.
The disappointing — be honest about these
This is where most wellness sites get it wrong: they take a hopeful early signal and present it as a proven benefit. We will tell you what failed.
Alzheimer's disease. Two large Phase 3 trials (EVOKE and EVOKE+, Cummings et al., The Lancet, January 2026) tested oral semaglutide in 3,800+ patients with early-stage Alzheimer's. After 104 weeks there was no significant difference in cognitive decline (CDR-SB) or activities of daily living between semaglutide and placebo. Novo Nordisk terminated its Alzheimer's programme in November 2025. Some biomarkers nudged in the right direction (~10% reduction in CSF p-tau181), but the clinical signal was absent. The honest summary: early hopes have not held up.
Parkinson's disease. Mixed. The Exenatide-PD3 Phase 3 trial (UK, 194 patients, 96 weeks) failed — no benefit on motor symptoms, non-motor symptoms, or brain imaging. The LIRA-PD trial of liraglutide showed a small, non-significant trend. Against that, LixiPark (lixisenatide, Phase 2, 156 patients, NEJM 2024) showed a positive motor-symptom signal: placebo got worse by 3.04 points on MDS-UPDRS while lixisenatide held essentially steady at −0.04 points (p=0.007). One promising Phase 2 trial against one failed Phase 3 trial is not a settled question. Until LixiPark's signal is confirmed in a larger Phase 3, GLP-1 drugs should not be used clinically for Parkinson's outside a trial.
Addiction, depression, PCOS, dementia prevention. Lots of small studies, lots of preprint hype, very little Phase 3 evidence. Treat with caution.
Side effects and risks — the honest list
| Risk | Frequency | Notes |
|---|---|---|
| Nausea, vomiting, diarrhoea | 30–50% in first three months | Usually mild; often fades with dose titration. |
| Pancreatitis | ~1% | Serious — requires medical attention immediately. |
| Gallbladder disease | ~5% | Sometimes needs surgery. |
| Thyroid C-cell tumours | Boxed warning | Avoid if you have a personal or family history of MEN-2 syndrome. |
| Diabetic retinopathy worsening | Uncommon | Monitor in diabetic patients. |
| Muscle loss | 15–40% of weight lost may be lean mass | Mitigable — see our companion piece on avoiding muscle loss. |
| Weight regain after stopping | Up to 2/3 within a year | Almost exclusively fat. Mitigable with diet + exercise transition. |
| Bone density loss | Emerging concern | Monitor in older patients. |
Counterfeit drugs — a public-safety warning
The FDA seized a batch of counterfeit Ozempic in December 2025 (lot PAR1229), distributed outside Novo Nordisk's authorised supply chain. The FBI issued companion safety warnings about counterfeit semaglutide through 2025. "Compounded" semaglutide and tirzepatide from non-licensed pharmacies has, on testing, contained the wrong ingredients, wrong doses, or no active ingredient at all. The FDA has proposed (public comment until June 2026) to formally prohibit 503B outsourcing facilities from compounding these drugs.
In South Africa, SAHPRA has not (as of May 2026) issued the same level of detailed counterfeit warnings as the FDA, but the risk applies. If you are considering a GLP-1 medication: see a registered doctor, get a script, and fill it at a registered pharmacy. Online "compounded semaglutide" from unknown sources has caused hospitalisations and deaths worldwide. This is the single most important safety message we can give you.
Where diet fits in
The clearest finding in the 2024–2026 obesity literature is that GLP-1 drugs work best as catalysts alongside lifestyle change, not as replacements for it. Three things drive the difference:
- Muscle preservation. A high-protein diet (1.2–1.6 g of protein per kg body weight per day) combined with resistance training preserves significantly more lean mass than either diet or drug alone.
- Weight maintenance after stopping. Patients who maintain the eating pattern they built up while on the drug retain most of their weight loss. Patients who treated the drug as a "magic bullet" and didn't change habits regain rapidly.
- Natural GLP-1 release. Fibre + protein together activate the same endogenous pathways the drugs mimic — by smaller margins, but real ones. For people who can't access or afford the drugs, the dietary lever is meaningful.
Our position: GLP-1 drugs are tools; how you eat is the system that holds the tool useful. We have a companion piece on eating like you're on Ozempic — without the drug, and another on how to avoid muscle loss while on it.
So should you take a GLP-1 drug?
That is not our question to answer — it is your prescribing doctor's. What we can say:
- If you have type 2 diabetes that isn't responding well to metformin alone, GLP-1 drugs are an evidence-backed second-line option that your endocrinologist will already be considering.
- If you have obesity (BMI ≥ 30, or ≥ 27 with co-morbidities), GLP-1 drugs are an evidence-backed option to discuss with your GP — alongside, not instead of, lifestyle change.
- If you are considering them for cosmetic weight loss without a medical indication, please understand: the side-effect profile is non-trivial, the cost is high, the rebound risk is real, and the long-term safety data is still accumulating. A doctor who will write that script knowing your full history is a very different proposition from an online clinic that asks five tickbox questions.
- If you are considering compounded or imported semaglutide because it's cheaper: don't. The counterfeit risk is real and people have died.
What we recommend whatever you decide
Eat like a person who values their body's metabolic future, whether or not you ever take one of these drugs. That means: lots of vegetables, enough protein to support your lean mass, fibre at every meal, and food you recognise as food. Our range of small-batch Karoo-grown ingredients is built around that premise. Our coach platform tailors a meal plan to your specific medical context, religion, culture, and what you actually like to eat — and explicitly defers to your doctor for medical advice. We are not a substitute for medical care. We are a complement to it.
References
- Wilding JPH et al. (STEP 1). N Engl J Med 2021;384:989–1002.
- Jastreboff AM et al. (SURMOUNT-1). N Engl J Med 2022;387:205–216.
- Lincoff AM et al. (SELECT). N Engl J Med 2023;389:2221–2232.
- Cummings J et al. (EVOKE / EVOKE+). The Lancet January 2026.
- Meissner WG et al. (LixiPark). N Engl J Med 2024;390:1176–1185.
- FDA Counterfeit Ozempic warning, December 2025; FDA proposed rule on 503B compounding 2026.
This article is general information, not medical advice. We do not diagnose, prescribe, or recommend starting, stopping, or changing the dose of any medication. Always consult a registered medical doctor for medical decisions.