Evidence-based articles on GLP-1 side effects, pricing, generics and safety.
Generic semaglutide is now launching in India, Canada, and Brazil. Is it the same as Ozempic? Here's what equivalence actually means — and what to look for.
Read moreOzempic costs €300+ per month in some countries and €14 in others. Here's the patent economics behind the price — and why 2026 is changing everything.
Read moreConsidering semaglutide for weight management? Here's what to expect in week 1, month 1, and beyond — from dose titration to side effects to lifestyle pairings.
Read moreCounterfeit and grey market semaglutide is a growing problem. Here's how to tell if what you're buying is the real thing — and what to do if you're not sure.
Read moreIndia's semaglutide patent expired March 20, 2026. Within 48 hours, 15 generic versions launched at ₹1,290/month — a 90% drop from Ozempic. Here's what it means for Indian patients.
Read moreA clear, non-hype explanation of what semaglutide actually does inside your body — the gut, the brain, the pancreas — and why that matters for how you take it.
Read moreThree brand names, two molecules, very different labels. Here's an honest comparison of Ozempic, Wegovy, and Mounjaro — what the differences actually mean for patients.
Read moreNausea is the most common reason people stop semaglutide in the first month. Here's a practical, evidence-based playbook for getting through the first four weeks.
Read moreGLP-1 drugs like semaglutide and tirzepatide are not for everyone. Here's an honest list of who should not take them, and why — from contraindications to caution cases.
Read moreGLP-1 drugs do a lot, but diet still matters — maybe more than you think. Here's what to eat to protect muscle, avoid nausea, and get the most out of semaglutide.
Read moreRetatrutide, an experimental triple-agonist weight loss drug, has been linked to emotional blunting and reduced romantic feelings. What does the science say?
Read moreFrom GLP-1 to dual and triple agonists: how do these molecules work in your body? The chemistry behind weight management, explained without jargon.
Read moreAn honest overview of the side effects, risks and safety data of weight management medication. What does the science say? And how do you minimise risks?
Read moreYou visit your dietitian but the weight won't budge. When is it time for medical support alongside nutritional advice? And what does it cost in 2026?
Read moreTried Weight Watchers, Noom or another diet programme without results? Discover why dieting alone often isn't enough and what medical alternatives are available.
Read moreBranded medications for weight loss cost €150-500/month and are mostly not reimbursed in 2026. Discover the alternative via compounding, priced independently with your pharmacy.
Read moreMedical weight loss doesn't have to be expensive. Compare the costs of every option in the Netherlands: from dietitians to compounding pharmacies. Discover the affordable alternative.
Read moreCompounding is the custom preparation of medication by a pharmacist. Discover how it works, why it's legal, and what the advantages are compared to standard branded products.
Read moreDiscover why more and more Dutch people are choosing personalised weight management programmes via compounding pharmacies. Affordable, effective and under medical supervision.
Read moreCompare compounded preparations with standard branded products. Costs, quality, dosage and safety — everything laid out for Dutch consumers.
Read moreThis article described a 4-step Magistra service — questionnaire, doctor assessment, pharmacy matching and delivery. That service does not exist. What a compounded treatment in the Netherlands generally requires, and what Magistra does publish: independent GLP-1 data.
Read moreHow safe are compounded preparations? An overview of Dutch and European regulation, quality controls and registration requirements for compounding pharmacies.
Read moreAn honest comparison of the costs of weight management programmes in the Netherlands. From dietitians to branded products to compounding — what does it really cost?
Read moreHow do you find a licensed compounding pharmacy in the Netherlands yourself, and what should you look for — BIG registration, IGJ inspection, facilities? Magistra does not operate a pharmacy finder; we publish independent GLP-1 data.
Read moreCalculate your BMI and discover when a personalised weight management programme makes sense. With an explanation of BMI categories and your options in the Netherlands.
Read moreMagistra stores no personal health record. What we do keep (waitlist email, contact message), where it lives (EU), and what GDPR rights you have.
Read moreDiet pills vs. compounding: what's the difference? Why personalised programmes via a doctor and pharmacy are safer and more effective.
Read moreAn evidence-grounded look at the Chinese semaglutide and tirzepatide peptide trade — what an independent lab test can and cannot tell you about a vial, the legal reality country by country, and what patients should know before relying on this supply chain.
Read moreAn evidence-grounded forecast of GLP-1 medication supply and demand through 2030. Manufacturing capacity, generic ramp, demand drivers (cardiovascular, kidney, sleep apnea, oral formulations), the role of insurance unlock, and what the 2031 OECD generic wave actually means for affordability.
Read moreGeneric semaglutide costs roughly $15-95 a month in India; a US patient pays $349 a month self-pay through Novo’s own pharmacy, and $1,349 at list only where their cost is list-linked. The trade that gap creates is narrower than the raw ratio suggests: it moves the list-price-exposed patient, not every cash payer. An evidence-grounded look at who travels, what gets through customs, the difference between the Indian-generic gray market and the Chinese-peptide gray market, and how the trade evolves to 2031.
Read moreSince February 2025 GPhC guidance has barred UK online pharmacies from prescribing Wegovy, Mounjaro and other weight-loss jabs on a questionnaire alone and required the prescriber to independently verify weight, height or BMI; in January 2026 the GPhC's revised inspection framework made failing that check an inspection failure. Here's what changed, why, and what it means for patients seeking treatment and doctors providing it.
Read moreWe audited every UK company with 'peptide' in its name on Companies House. 98 are brand new with no track record, none of the credible ones supply GLP-1, and one woman has already died. Here's what patients need to know — and the safe alternative.
Read moreWe track 15 GLP-1 side effects in our real-world evidence database. After a 31 August 2026 correction excluded seed rows from the eligible base, only 1 cleared our bar for a fully calibrated statistical model; as the corpus grew, 13 of the 15 clear it as of 30 September 2026. Here's which ones, why, and what's published instead for the rest.
Read more14 of the 15 GLP-1 side effects we track have a published rate pooled directly from our own corpus of clinical trial and regulatory sources — not copied from a single paper. Here is every number, with its confidence interval, exactly how many sources it rests on, and the corrections that shaped the eligible base.
Read moreOf the distinct community reports we hold from people on GLP-1 medications — Reddit threads and Drugs.com patient reviews — which side effects come up most? All 15 reporting frequencies with their denominator, what this number deliberately is not (an incidence rate), why we count mentions instead of averaging what posts claim, the 2026-08-29 correction that cut the pool from 185 reports to 26, and the 2026-09-25 restatement that admitted ten GLP-1 communities and took it to 42.
Read moreThe US label warns that gallstone risk goes up. A meta-analysis of 76 randomized trials agrees — especially at weight-loss doses. A 32,052-patient cohort study published in December 2025 found the opposite direction in one high-risk population. Here is every number, the population it belongs to, and what our own evidence base holds.
Read morePersonalised risk tools give you one number. It is really two ingredients multiplied together — an evidenced base rate and a set of profile multipliers — and almost nobody shows you the second. Here are all 75 of ours, including the fact that none of them carries a citation.
Read moreThe US semaglutide label puts hair loss at 3% versus 1% on placebo, and says it was associated with weight reduction and reported several times more often in women than in men. The tirzepatide registry posts alopecia per arm at about 5% versus 1.1%, and as of 7 September 2026 those three arms are also our own database's first eligible clinical rates for the effect — here is every number, the arm and n it belongs to.
Read moreSearch "how common is vomiting on Ozempic" and every answer gives you one number with no source. Our own corpus held 52 distinct sources that state a vomiting rate on 30 September 2026, and they range from 0.0% (a 30-person drug-interaction study) to 45% (a 10-person early-phase study), with the 8,803-patient SELECT cardiovascular trial at 6.5%. Here is a fixed sample spanning the size range, its population, its n, our pooled 10.6%, and why the range is this wide rather than a defect in the data.
Read moreThe US semaglutide label puts constipation at 24% on Wegovy 2.4 mg against 11% on placebo, so nearly half the headline is the background rate. We read the placebo arm of every constipation source in our corpus that posts one — from an 8,803-patient cardiovascular trial to a 61-participant dose-ranging study — and in the weight-reduction trials the label itself pools, the drug-minus-placebo difference is 11 to 14 points every time. Here is every arm, its n, and why our own pooled figure, about half the label's, is not the Wegovy number.
Read moreThe US semaglutide label puts headache at 14% on Wegovy 2.4 mg against 10% on placebo — the closest drug-to-placebo pair in its table. We read the placebo arm of every headache source in our corpus that posts one — STEP 1 15.2% against 12.2%, SURMOUNT-1 7.0–8.0% against 7.3%, three orforglipron Phase 3 trials within two points of placebo — and in every arm of more than 300 people the drug sits within about three points of its own placebo. Here is every arm, its n, the two large trials where headache never reached the reporting threshold at all, and the one row we withheld before publishing.
Read moreThe US semaglutide label puts abdominal pain at 20% on Wegovy 2.4 mg against 10% on placebo — but that row groups seven adverse-event terms, and STEP 1, one of the trials it pools, posts the single term on ClinicalTrials.gov at 9.7% against 5.5%. We read the placebo arm of every abdominal pain source in our corpus that posts one: in every arm of more than 300 people the drug adds 1.5 to 4.2 points for the single term.
Read moreThe US semaglutide label puts nausea at 44% on Wegovy 2.4 mg against 16% on placebo — and unlike its abdominal pain, fatigue and gastritis rows, the nausea row groups no other terms, so the three trials behind it sum to 44.8% and 15.5% in their own registry records. But the same molecule at the same dose reads 18.1% in SELECT, the largest trial of the class, and the placebo arms of twelve placebo-controlled trials run from 0% to 25%. We re-read every arm of all 28 sources in our nausea base against the live registry before quoting one — and found that a trial behind the label's own number was missing from it.
Read moreThe US semaglutide label puts diarrhoea at 30% on Wegovy 2.4 mg against 16% on placebo — the joint-highest placebo figure in its table — and unlike the abdominal pain row it groups no other adverse-event terms, so the three trials it pools sum to 30.4% and 15.6% in their own registry records. We re-read all 21 registry records in our diarrhoea base against the live registry: of 25 drug arms, 24 report more diarrhoea than their own placebo arm, and every drug arm of more than 300 people sits 1.4 to 3.0 times its own placebo. But the same molecule at the same maintenance dose reads 31.5% in STEP 1 and 10.6% in SELECT, whose registry record says its adverse events were not all collected systematically.
Read moreThe US semaglutide label puts fatigue at 11% on Wegovy 2.4 mg against 5% on placebo — but that row groups two adverse-event terms, fatigue and asthenia, which trial registries post separately and we count separately. On 18 September 2026 we read the placebo arm of all seven sources then in our fatigue base that post one of at least 20 people: every drug arm of more than 300 people sat 0.2 to 6.0 points above its own placebo, and in ACHIEVE-1 all three orforglipron doses report less fatigue than placebo. We also reported the thing a reader of our own number most needs to know: 38 of the 65 eligible stated rates we then held came from a single Phase 1 bioequivalence study in otherwise-healthy volunteers.
Read moreThe US semaglutide label puts gastro-oesophageal reflux at 5% on Wegovy 2.4 mg against 3% on placebo — three of those five points were also reported by people on no drug, a larger placebo share than nausea, vomiting, diarrhoea or constipation. We read the placebo arm of all eight sources in our acid-reflux base that post one of at least 20 people: every placebo-controlled drug arm of more than 300 people sits 1.9 to 4.1 points above its own placebo. Also disclosed: the same label table reports dyspepsia at 9% and belching at 7% — both higher than reflux and both with far wider margins over placebo — and neither term is in our base, which counts one registry term only.
Read moreThe US semaglutide label puts dizziness at 8% on Wegovy 2.4 mg against 4% on placebo — half of the dizziness reported on the drug was also reported on placebo. On 18 September 2026 we read the placebo arm of all ten sources then in our dizziness base that post one of at least 20 people: every placebo-controlled drug arm of more than 300 people sat 0.4 to 7.4 points above its own placebo, three small arms sat below it, and the two largest arms in our base — 6,647 people each on tirzepatide and on dulaglutide — read 7.0% and 6.1% with no placebo to compare against. Also disclosed: when this was written, 59 of our 97 eligible stated rates came from two Phase 1 orforglipron studies in healthy volunteers.
Read moreThe US semaglutide label uses the word appetite three times and all three are in the pharmacology sections — it is the mechanism of action, and it appears in no adverse-reaction table at all. But the three trials the label pools identify themselves by their arm sizes, and their own registry records post decreased appetite for 197 of 2,116 people on the drug (9.3%) against 44 of 1,261 on placebo (3.5%) — a margin that would have ranked it ninth of twenty rows in the label's own table, above headache and dizziness. Across the 23 studies in our base when it was written, the reported rate ran from 0.6% to 70%, the placebo arms from 0% to 50%, and a Phase 4 trial designed to measure appetite reports one of the lowest rates of all.
Read moreAsk how common nausea is on a GLP-1 and you get one number pooled across molecules. Ours, when this was written on 19 September 2026, was 23.0% from 135 stated rates across 29 studies — and 103 of those 135 rows (76%) came from one drug, orforglipron, which moved 6.0% of the figure, while semaglutide's 16 rows moved 57.6% of it. That divergence held in 12 of our 14 pooled estimates then, and still does on 30 September 2026. From 19 September 2026 the composition is published as a field on our API rather than left for a reader to derive: which molecules the rows describe, and what share of the weighted mean each one actually carries.
Read moreFour numbers describe GLP-1 pancreatitis and no two share a denominator: our pooled 0.1% (11 stated rates, 3 studies), the FDA label's 0.2 adjudicated cases per 100 patient-years, a 4.5% share of liraglutide's most-reported FAERS terms, and the one nobody publishes — the placebo arm. We read every pancreatitis term in the results tables of both placebo-controlled trials in our base: SELECT posts 27 affected participants on placebo against 20 on semaglutide, SURMOUNT-1 posts 0.31% on placebo against 0.21% across three tirzepatide arms. Our own 0.1% is a drug-arm figure covering two of the registry's six terms, and at this event rate it is not distinguishable from what the same trials report on no drug. Update, 4 October 2026: a third placebo-controlled trial, SURMOUNT-OSA, has since entered our base and posts acute pancreatitis in 1 of 233 on tirzepatide against 0 of 234 on placebo, also inside chance.
Read moreNausea, our most-evidenced side effect (135 stated rates, 29 distinct sources), publishes a 95% interval of 6–58%. Hair loss, one of our least-evidenced (8 rates, 3 sources), publishes 2–12% — visibly tighter. A reader skimming ranges alone would trust the narrow one more. Our own numbers say the opposite is closer to true, and the reason is a specific, checkable statistical mechanism, not a house quirk.
Read moreOur confidence grade is a bucket on distinct source count, computed without ever reading the rate itself. As of 30 September 2026, eight effects share our top 'very_high' grade — and pool to anywhere from 6.2% to 20.6%, a 3.3x gap. At publication (20 September) it was diarrhoea and acid reflux sharing 'high' at a 3.4x gap; the corpus has since grown past that pairing, but not past the mismatch itself.
Read more'Ozempic personality' has been covered by the Washington Post, the Guardian, the BBC, MedPage Today and CNBC — 20 news-typed rows sit in our corpus. But of our 15 tracked side effects, emotional blunting is the only one with zero corpus-derived clinical rate — and until 25 September 2026 it had zero community mentions too (one since, on a label we cannot re-read). Not a rounding error: a specific, checkable gap in what evidence exists to measure.
Read moreOur evidence base for GLP-1 injection-site reactions is four stated rates from two clinical trials — three tirzepatide dose arms from SURMOUNT-1, and one seven-person dulaglutide arm from a kidney-disease trial. The pooled estimate is 4.5%, with a 95% confidence interval spanning 0% to 53%. Here is every number, why the interval is that wide, and what FDA's spontaneous-report data shows instead.
Read moreWe ran an arithmetic test on our own evidence base: sum the at-risk denominators of the trial arms we store, and if they exceed the trial's own enrolment, some participants are inside more than one of our rows. On 21 September 2026, six of 27 registry studies failed it, and 726 of our then 1,134 registry rate rows came from those six. Removing all six moved every published pooled rate by at most 0.5 percentage points — and moved three source-diversity grades down a step. Here is the full before-and-after table.
Read moreWe measured the dose gradient the only fair way posted trial results allow: between arms of the same trial, on the same drug. Across 78 comparisons from 8 trials, going from a trial's lowest dose arm to its highest adds a median of 2.7 percentage points to gut effects and 0.8 points to everything else — far less than the framing implies. It also shows that the 23%-to-42% high-dose jump on our own predictor is mostly a change of drug mix, not of dose. And of our 1,142 eligible rates, zero describe a starting dose. (Figures as of 22 September 2026; restated 29 September: 162 comparisons across 19 trials, gut effects +3.0 points, everything else +1.2, and 21 rates now sit on a starting dose — see the dated restatement in the article.)
Read moreAlmost every “tirzepatide vs semaglutide side effects” comparison splices together rates from different trials with different populations and durations — the same confound our dose-gradient piece found for dose. One trial in our corpus avoids it with both arms' registry counts: SURMOUNT-5 (NCT05822830) randomized 751 adults between the two drugs, titrated to each one's 15 mg or 2.4 mg maintenance dose, in the same protocol. Of 12 shared effects, tirzepatide's rate is lower for 7, tied for 1, and higher for 4 — including an 8.6%-vs-0.3% injection-site-reaction gap that a two-proportion test says is not noise, and two more gaps that only look that way. These are one trial's registry rows, not our pooled corpus estimates.
Read moreADJUNCT ONE (NCT01836523) randomized 1,398 adults with type 1 diabetes to liraglutide 0.6, 1.2 or 1.8 mg, or placebo, added to insulin. Three clean dose arms, no cross-trial confound, and a clear gut-effect gradient: nausea 32.0% → 40.5% → 49.6%, vomiting 6.9% → 18.4%, reduced appetite 8.9% → 18.4%. But this is a type 1 diabetes population on background insulin, not the weight-management population our predictor is built for — and the trial's own published safety signal was hypoglycemia and ketosis, neither of which our 15 tracked effects cover.
Read moreAsk an AI what happens when you stop Ozempic or Wegovy and you'll hear "you regain the weight." Two semaglutide trials actually measured it. The STEP-1 extension followed 327 trial completers for a year after everyone stopped: two-thirds of the lost weight came back. STEP-4 randomized 803 people to keep taking the drug or switch to placebo for 48 weeks, and measured from baseline about half came back. That trial-measured two-thirds is the only evidenced part of our discontinuation tool — the five adjustments it applies on top are our own judgment, and the tool does not vary by drug because no trial in it measured that.
Read moreAsk an AI what happens if you stop Mounjaro or Zepbound and you may hear "you regain two-thirds of the weight." One randomized trial measured it. SURMOUNT-4 took 670 people who had lost 20.9% on tirzepatide and randomized them to continue or switch to placebo for 52 weeks. The placebo group's weight rose 14% from where it stood at randomization — which, measured from their starting weight, is about half of the loss coming back, not two-thirds. Here is the arithmetic, the like-for-like semaglutide comparison from STEP-4, and what our own tool does and does not do with it.
Read moreMost semaglutide-vs-dulaglutide side-effect comparisons splice rates from different trials. SUSTAIN 7 (NCT02648204) randomized 1,201 adults with type 2 diabetes between the two drugs at two dose levels for 40 weeks. At the low pair (0.5 vs 0.75 mg) semaglutide reported more of all six listed effects, nausea 22.3% vs 13.0% the only gap clearing a 12-test correction; at the high pair (1.0 vs 1.5 mg) the drugs sat within 1.4 points on five of six, with dulaglutide higher on diarrhoea. Every row read from the registry.
Read moreACHIEVE-3 (NCT06045221) randomized 1,698 adults with type 2 diabetes on metformin between two daily GLP-1 tablets, orforglipron and oral semaglutide, at two dose levels for 52 weeks. At the low pair (12 vs 7 mg) orforglipron reported more nausea (26.4% vs 13.1%), vomiting (17.0% vs 7.7%) and diarrhoea (21.7% vs 13.1%), all three clearing a 16-test correction; at the high pair (36 vs 14 mg) no gap cleared it. Open-label and Lilly-funded. Every row read from the registry.
Read moreSURPASS-2 (NCT03987919) randomized 1,879 adults with type 2 diabetes on metformin to tirzepatide 5, 10 or 15 mg or semaglutide 1 mg, open-label, for up to 44 weeks. Across six tracked side effects and three tirzepatide doses — 18 comparisons — the largest gaps were diarrhoea at 10 mg (+4.9 points, z=2.17) and reduced appetite at 15 mg (+3.6 points, z=2.15), both past the single-test line and short of the Bonferroni-corrected bar. No gap is established; two are suggestive. Every row read from the registry.
Read moreThe trials behind our pooled rates tracked adverse events for anywhere from 1.7 to 364 weeks — 12 days to 7 years. As of 3 October 2026, 755 of the 1,671 eligible registry rows that state a counting window (45.2%) come from trials of 26 weeks or less, yet across seven tracked effects those short trials carry 1.1–2.4% of the pooled weight, not their row-count share. Nausea's own median stated rate is 8.0% in trials of 26 weeks or less and 25.2% in trials over two years — computed by the same windowMix function that runs on magistra.health, and checked against the live data store.
Read moreSee your personal GLP-1 side-effect risk