Ozempic-class drugs

Semaglutide (Ozempic, Wegovy, Rybelsus) and the next-generation dual agonist tirzepatide (Mounjaro, Zepbound) are the first drugs in any class to combine double-digit weight loss, 20% reduction in major cardiovascular events, and measurable deceleration of validated epigenetic aging clocks, all in non-diabetic adults. That combination is genuinely new in pharmacology. Whether they should be used by healthy adults purely for longevity — and at what dose — is the central question this article tries to answer honestly.

What they are, briefly

GLP-1 (glucagon-like peptide-1) is an incretin hormone gut L-cells release after a meal. It boosts glucose-dependent insulin secretion, suppresses glucagon, slows gastric emptying, and acts centrally to dampen appetite. Native GLP-1 has a half-life of minutes. Semaglutide is a structurally modified analog resistant to degradation by the enzyme DPP-4 (dipeptidyl peptidase-4), allowing once-weekly subcutaneous (Ozempic, Wegovy) or daily oral (Rybelsus) dosing. Tirzepatide adds GIP (glucose-dependent insulinotropic polypeptide) receptor agonism on top.

Efficacy in non-diabetic adults

Weight loss — Strong

In non-diabetic adults with obesity, semaglutide 2.4 mg/week produces 10–16% body weight loss over 68–104 weeks. A 2026 meta-analysis of 7 randomised trials in 5,411 non-diabetic adults found a mean difference of −12.24 kg, −12.15% body weight, and −9.32 cm waist circumference versus placebo, with patients 2.6× more likely to hit clinically meaningful weight thresholds.^[1]

The 4-year SELECT extension tracked >17,600 non-diabetics with overweight/obesity and established cardiovascular disease (CVD); participants sustained an average 10.2% weight loss and 7+ cm waist reduction at 208 weeks. 12% achieved a normal BMI vs. 1% on placebo.^[2] Oral semaglutide 25 mg in OASIS 4 produced 13.6% weight loss vs. 2.2% placebo over 64 weeks in non-diabetic obesity.^[3]

Tirzepatide is meaningfully more effective. In the head-to-head SURMOUNT-5 trial — an open-label randomized comparison in adults with obesity and no diabetes — tirzepatide produced 20.2% weight loss at 72 weeks vs. 13.7% for semaglutide (waist 18.4 vs. 13.0 cm), a 6.5-point gap that works out to roughly 47% greater relative weight loss.^[4] The endpoint here is weight, a surrogate rather than a hard clinical outcome, and the trial was funded by the manufacturer, Eli Lilly.

Outside the obesity indication — Weak / no controlled evidence

There is no randomised evidence for semaglutide in adults with a BMI under 27 — every registration trial (STEP, SURMOUNT, SELECT) required overweight or obesity to enroll. What circulates instead are uncontrolled retrospective chart reviews from private clinics dispensing compounded semaglutide, published in low-tier journals; they report that most patients lose weight, which is what an appetite suppressant does and tells you nothing about benefit or harm. This kind of "elective" use sits outside the FDA label and remains regulatorily and ethically contested — the absence of a medical necessity changes the risk-benefit math sharply, especially for muscle and bone (see below).

The geroprotective signal — Weak / preliminary

This is what makes GLP-1 receptor agonists (RAs) interesting on a longevity site rather than just a weight-loss site. The mechanistic case has been spelled out across all twelve recognized hallmarks of aging: GLP-1 signaling improves insulin sensitivity (modulating mTOR), activates AMPK and SIRT1 (the catabolic counterweights), promotes mitochondrial biogenesis, restores autophagy, reduces senescence markers, dampens "inflammaging" — chronic low-grade inflammation — lowering interleukin-6 (IL-6), tumour necrosis factor-α (TNF-α), and high-sensitivity C-reactive protein (hsCRP) independent of weight loss, and improves endothelial nitric oxide production.^[5]

The first human RCT to test this with validated biomarkers used adults with HIV-associated lipohypertrophy — a population that ages biologically faster than the general population. After 32 weeks of weekly semaglutide:

^[6] Organ-specific analyses showed concordant deceleration across 11 biological systems, strongest in the inflammatory, brain, and cardiac compartments.

Read those numbers with the design in mind. This was a post-hoc, exploratory analysis of a 32-week phase 2b trial — not a trial designed to test aging clocks — with 45 participants on semaglutide and 39 on placebo, in a population selected for accelerated biological aging, and the epigenetic profiling was performed by a commercial clock vendor whose product is the outcome measure. It is also a preprint that has not yet been peer-reviewed, and the authors themselves ask for validation in broader populations.

With that said: it is the first time a drug has moved multiple validated molecular aging biomarkers in a randomized human trial. It does not prove longevity extension — none of these clocks have been validated as causal, and caloric restriction moved DunedinPACE in the general population two years earlier (CALERIE). It places GLP-1 RAs alongside, not above, the rest of the geroprotector class on translational evidence. Compare rapamycin and metformin in Geroprotectors: mechanism rich, hard-outcome evidence still pending.

Cardiovascular and functional benefits — Strong (cardiovascular events); Moderate (functional capacity)

The SELECT trial is the hard-outcome anchor for the entire class: in about 17,600 non-diabetic adults with overweight or obesity and established cardiovascular disease, semaglutide 2.4 mg cut major adverse cardiovascular events — MACE, meaning cardiovascular death, non-fatal myocardial infarction (MI), and non-fatal stroke — by roughly 20% over a mean ~40 months, with events in 6.5% on semaglutide versus 8.0% on placebo (hazard ratio 0.80, 95% confidence interval 0.72–0.90 — the confidence interval is the range where the true effect most plausibly lies, and because it stays below 1.0 here the benefit is unlikely to be a chance finding).^[7] This is a randomized trial reporting a hard clinical outcome — the strongest tier of evidence — though the trial was funded by the manufacturer, Novo Nordisk. Real-world claims data corroborate, and the cardioprotective effect appears to be independent of the magnitude of weight loss — implicating direct vascular and myocardial mechanisms.

Functional capacity also improves. In STEP-HFpEF, patients with obesity-related heart failure with preserved ejection fraction gained a mean +15.1 m on the 6-minute walk test on semaglutide.^[8] The phase 3b STRIDE trial randomised 792 adults with type 2 diabetes and symptomatic peripheral artery disease with intermittent claudication to semaglutide 1 mg weekly or placebo for 52 weeks; maximum walking distance on a constant-load treadmill improved by an estimated treatment ratio of 1.13 over placebo (1.21× vs 1.08× baseline, p = 0.0004) — roughly 40 additional metres.^[9] Both signals are clinically meaningful in populations that are very hard to move with anything else — though note STRIDE enrolled only diabetic patients, so it does not generalise to non-diabetic claudication.

The brain: food noise, mood, and addiction

The most psychologically striking effect patients report is the silencing of "food noise" — the relentless, intrusive cognitive preoccupation with food. Mechanistically this maps to GLP-1 receptors in the ventral tegmental area (VTA) and the brain's default mode network: semaglutide blunts the dopamine spike from hyperpalatable food, which extinguishes the cue-driven, impulsive reinforcement loop.

The blunting generalises beyond food (Moderate — large, but observational). A Swedish national register study followed 95,490 people who already carried a diagnosis of depression or anxiety and were taking antidiabetic medication, comparing each person's own periods on and off a drug. Against their own non-use periods, semaglutide was associated with a 44% lower risk of worsening depression (aHR 0.56, 95% CI 0.44–0.71), a 38% lower risk of worsening anxiety (0.62, 0.52–0.73), and a 47% lower risk of worsening substance use disorder (0.53, 0.35–0.80); GLP-1 RAs as a class were associated with less self-harm (0.56, 0.34–0.92).^[10] Two limits on how far that stretches: the outcome is deterioration (psychiatric hospitalisation, long psychiatric sick leave, self-harm, suicide) in people who were already ill, not prevention of new depression in healthy adults; and the effect was specific to semaglutide and liraglutide — exenatide and dulaglutide showed nothing. Patients also spontaneously report reduced cravings for alcohol, opioids, nicotine, and even compulsive shopping. An NIH-funded RCT showed adding weekly low-dose semaglutide to cognitive behavioural therapy further reduced heavy drinking.^[11] See also Alcohol for context on alcohol harms.

Caveat: profound dopamine modulation can theoretically blunt non-food pleasure (anhedonia) in a susceptible minority. Watch for it on initiation and if mood drops, treat it as a real signal.

Alzheimer's: biomarkers moved, clinical decline didn't — Strong (null for clinical benefit)

The phase 3 EVOKE / EVOKE+ trials (n>3,800, oral semaglutide, ~3 years) failed the primary cognitive endpoint — semaglutide did not slow clinical progression from mild cognitive impairment to mild Alzheimer's vs. placebo.^[12] Notably, cerebrospinal fluid (CSF) biomarkers did move in the right direction — pTau181 −8.3%, total tau −6.6%, neurogranin −8.6% relative to slight rises on placebo — and CSF proteomics shifted away from neuroinflammatory signatures. The honest interpretation: GLP-1s may modify pathology biology without rescuing already-established clinical disease, leaving open the prophylactic-decades-earlier hypothesis but offering nothing for current symptomatic patients.

The catch nobody talks about enough: muscle and bone

This is where elective use among healthy adults diverges sharply from cost-benefit analysis in obesity treatment.

Up to 40% of the weight is lean mass — Moderate

Across studies, roughly 25–40% of total weight lost on semaglutide originates from fat-free mass — skeletal muscle, connective tissue, and water. Older adults and women are most susceptible. Muscle is the largest sink for postprandial glucose; losing it directly worsens insulin resistance, lowers resting energy expenditure, and accelerates frailty risk. Worse, if the drug is later discontinued, weight rebounds almost entirely as fat — at a permanently lower metabolic baseline.

The BELIEVE trial tested combining semaglutide with bimagrumab, an anti-myostatin monoclonal antibody. Over 72 weeks, the combination produced 22.1% total weight loss, of which 92.8% was fat mass. Semaglutide alone matched the dose lost 15.7% with only 71.8% from fat — a substantial muscle hit. Bimagrumab alone lost 10.8% (all fat) while adding 2.5% lean mass.^[13] Anti-myostatin drugs aren't approved yet, so for now the answer is mechanical (resistance training) and nutritional (protein) — see below.

A nuance: while muscle quantity drops, muscle quality in the remaining tissue may improve, with reduced myosteatosis (intramuscular fat), preserved mitochondrial respiration, and better insulin sensitivity in the residual fibres. This does not rescue absolute strength or sarcopenia risk, but it explains why functional outcomes (6MWT, HFpEF symptoms) improve despite lean mass loss.

Bone mineral density — Moderate

A 52-week phase 2 randomised trial in adults at increased fracture risk found semaglutide 1.0 mg reduced hip BMD by ~2.6%, lumbar spine BMD by ~2.1%, and tibial cortical thickness by 1.8% versus placebo, alongside ~9% weight loss.^[14] Most of the bone loss is driven by mechanical unloading and rapid energy deficit, not direct osteoblast toxicity. A matched-cohort scan study places the signal more in non-diabetics losing weight on GLP-1s than in people with type 2 diabetes: total-hip loss exceeded matched controls only in the non-diabetic group (−1.0% vs −0.6%), while in type 2 diabetes it was no different from controls — the metabolic improvements appear to offset the mechanical unloading.^[15]

A ~2.6% hip BMD drop over a year is clinically meaningful, especially in postmenopausal women and older adults. Baseline dual-energy X-ray absorptiometry (DXA), calcium and vitamin D repletion, and weight-bearing/resistance loading are not optional.

How to take it without wrecking yourself

If you are going to use a GLP-1 RA — especially as a healthy adult electing it for cardiometabolic optimization rather than treating obesity — the lifestyle scaffolding is the difference between a longevity tool and an accelerated path to sarcopenic obesity.

Protein

Joint guidance from the American College of Lifestyle Medicine, American Society for Nutrition, Obesity Medicine Association, and The Obesity Society sets the daily target at 1.6–2.2 g/kg body weight.^[16] For a 70 kg adult that's 110–155 g protein/day, distributed across 3–4 meals at ~30–40 g per serving (the leucine threshold for muscle protein synthesis).

The hard part: GLP-1s blunt appetite and slow gastric emptying enough that hitting these targets from whole food alone is genuinely difficult for most users. Whey protein (high leucine) or a complete plant blend (pea + rice) is essentially mandatory. The ongoing LEAN-PREP trial is using MRI quad cross-sectional area to validate the protein + resistance training protocol that prevents muscle loss on GLP-1 therapy.^[17] See also Protein for the underlying physiology.

Resistance training

Mandatory, not optional. 2–4 progressive resistance sessions per week targeting all major muscle groups. Without mechanical tension, even a high-protein diet will not halt muscle catabolism in a severe drug-induced caloric deficit. See Resistance training.

Micronutrients and fiber

A 16–39% drop in caloric intake means proportional drops in micronutrient ingestion. A narrative review pooling six studies in 480,825 adults on GLP-1 RAs found new nutritional deficiency diagnoses in 12.7% of users by 6 months, most often vitamin D (7.5% at 6 months, rising to 13.6% at 12 months), followed by nutritional anaemia (4%), iron deficiency (3.2%), and B-vitamin deficiency (2.6%); case reports of severe thiamine deficiency, including Wernicke encephalopathy, exist at the tail.^[18] Highest-priority adds:

• Vitamin D and calcium — vitamin D deficiency in 13.6% of users at one year; critical for the bone density side.
• Vitamin B12 and iron — altered gastric emptying and reduced intake; ~4% develop nutritional anemia.
• Vitamins A, E, and magnesium — reduced fat-soluble vitamin intake on low-fat anti-nausea diets.
• Fiber 22–34 g/day — counteracts severe constipation from delayed gastric emptying. Titrate up slowly with water.

A Mediterranean-pattern baseline (Dietary patterns) handles much of this without explicit supplementation.

Hydration and electrolytes

GLP-1s blunt central thirst signaling, which is exactly where chronic under-hydration causes the most damage long-term. Target ~2.5–3.5 L/day (91–125 oz) of properly mineralized water — see Water for the filtration and remineralization side. Add electrolytes (sodium, potassium, magnesium) on training days or with persistent GI side effects; this is one of the few standard adult contexts where daily electrolyte supplementation is genuinely indicated.

Bone surveillance

Baseline DXA before initiating, repeat at 12 months, then per risk profile. Vitamin D 25(OH)D ≥75 nmol/L; calcium 1,000–1,200 mg/day from food + supplement if needed; bone turnover markers (CTX, P1NP) for high-risk patients.

Dosing, microdosing, and the maintenance question

Standard titration

Wegovy titrates monthly: 0.25 → 0.5 → 1.0 → 1.7 → 2.4 mg/week. Slower titration reduces nausea and improves adherence.

"Microdosing" for longevity

A growing off-label practice among biohackers and longevity-focused clinicians is to hold at 0.25 mg/week or less indefinitely, never escalating. The argument: most of the systemic geroprotective effects (lower hsCRP, improved lipids, glycemic stabilization, reduced food noise, vascular insulin sensitization) plausibly occur at sub-weight-loss doses, while side effects, muscle loss, and bone loss scale with dose and weight loss magnitude.^[19]

Honest assessment of microdosing as of 2026:

• No RCT has tested low-dose chronic semaglutide for longevity endpoints in healthy adults.
• The epigenetic clock data above came from full doses in a non-healthy population — extrapolating to healthy adults at 10–25% of that dose is speculation.
• Microdosing has driven explosive demand for compounded semaglutide, which bypasses FDA oversight on dose accuracy, sterility, and salt formulation. Several state and federal warnings have been issued.^[20]
• Real risk: at sub-therapeutic plasma concentrations, you may capture none of the supposed benefits and still face the GI, gallbladder, and pancreatitis risk profile.

If you go this route, do it with a real prescriber, not a compounding mill, and keep DXA + protein + training in place anyway.

You probably can't ever stop

The randomized withdrawal evidence is unambiguous. In the STEP 1 extension trial, patients switched to placebo after reaching goal regained roughly two-thirds of lost weight (~11.6% body mass) within 68 weeks, and most cardiometabolic improvements reverted toward baseline.^[21] Like statin therapy and antihypertensives, semaglutide treats a chronic, relapsing condition; benefits last only while the drug is on board. Realistic expectation: indefinite maintenance, ideally titrated down to the lowest dose that holds the result (often 1.0–1.7 mg).

Cautions and interactions

The tolerability profile is not incidental to the decision — it is most of the reason people stop, and a few of the signals are serious. Incidence figures below come from the randomised trials and from post-marketing pharmacoepidemiology.

• Gastrointestinal adverse events — the near-universal cost. In the pooled STEP 1–3 analysis of semaglutide 2.4 mg vs placebo, nausea affected 43.9% vs 16.1%, diarrhoea 29.7% vs 15.9%, vomiting 24.5% vs 6.3%, and constipation 24.2% vs 11.1%. The great majority were mild-to-moderate, transient, and clustered around dose escalation; 4.3% of the semaglutide arm discontinued permanently because of them, versus 0.7% on placebo.^[22] Slower titration is the main mitigation.
• Gallbladder and biliary disease. A meta-analysis of 76 randomised trials in 103,371 patients found GLP-1 RAs raised the risk of gallbladder or biliary disease by about 37% (RR 1.37, 95% CI 1.23–1.52), driven by cholelithiasis and cholecystitis. The risk was more than doubled in trials using GLP-1 RAs for weight loss specifically (RR 2.29, 1.64–3.18) and scaled with dose and duration. Rapid weight loss is itself lithogenic, so this is partly the weight loss and partly the drug.^[23]
• Pancreatitis, gastroparesis, bowel obstruction. In a claims analysis of adults using GLP-1 RAs for weight loss, incidence per 1,000 person-years on semaglutide was 4.6 for pancreatitis and 9.1 for gastroparesis. Versus the comparator weight-loss drug bupropion-naltrexone, GLP-1 RAs carried a hazard ratio of 9.09 (1.25–66.00) for pancreatitis — roughly a 9-fold relative increase — about 4-fold (4.22, 1.02–17.40) for bowel obstruction, and about 3.7-fold (3.67, 1.15–11.90) for gastroparesis. The absolute risks are low; the relative increases are not, and the confidence intervals are wide.^[24] A prior episode of pancreatitis is a reason not to start.
• Sudden vision loss (NAION). A 2024 JAMA Ophthalmology matched cohort at a single neuro-ophthalmology centre found semaglutide associated with non-arteritic anterior ischemic optic neuropathy — a painless, largely irreversible loss of vision in one eye from optic-nerve infarction. In patients with type 2 diabetes, 36-month cumulative NAION incidence was 8.9% on semaglutide vs 1.8% on non-GLP-1 comparators (HR 4.28, 95% CI 1.62–11.29), with risk highest in the first year.^[25] That cohort was a referral population and the incidence is far above the population base rate, so the absolute risk for a typical user is much smaller — but a much larger cohort of 174,584 matched pairs with diabetes replicated the direction, with roughly 2.4× risk at 2–3 years.^[26] Any new painless visual loss on a GLP-1 warrants same-day ophthalmology, and the drug should not be restarted.
• Muscle and bone loss. Covered above. 25–40% of weight lost is fat-free mass; hip BMD falls ~2.6% over 52 weeks. This is the caution most relevant to a healthy adult using the drug electively.
• Contraindications. Personal or family history of medullary thyroid carcinoma (MTC) or multiple endocrine neoplasia type 2 (MEN2) — the rodent C-cell tumour signal — prior severe pancreatitis, pregnancy or planned pregnancy, and active eating disorders. Frail older adults at high sarcopenia risk should be screened before, not after.
• Drug interactions. Delayed gastric emptying alters the absorption of orally administered drugs — clinically relevant for narrow-therapeutic-index agents and for oral contraceptives around dose escalation. In insulin-treated or sulfonylurea-treated patients, doses need reducing to avoid hypoglycaemia. Anaesthetists now routinely ask about GLP-1 use because of retained gastric contents and aspiration risk; expect to be told to hold doses before elective surgery.
• Compounded semaglutide. Not FDA-reviewed for dose accuracy, sterility, or salt form. Dosing errors from patient self-drawing of unlabelled vials have driven poison-centre calls. If you use the drug, use the licensed product.

What we don't know yet

• Hard longevity outcomes in healthy adults. Every long-term outcome trial (SELECT, STEP-HFpEF, STRIDE) was in disease populations. No RCT has tested whether GLP-1s extend healthspan in metabolically healthy adults.
• Decades-long safety. Pancreatitis, gallbladder pathology, the rodent medullary thyroid signal, gastroparesis, and the NAION vision signal are all documented (see Cautions). Multi-decade exposure data don't exist yet.
• Microdosing dose-response. Plausible but unproven.
• Who shouldn't take it. Pregnancy, MEN2/MTC family history, history of severe pancreatitis, and active eating disorders are clear contraindications. Frail older adults at high sarcopenia risk should be screened carefully and only treated with aggressive scaffolding.

Bottom line for a healthy midlife adult

For a healthy adult at a normal BMI seeking longevity benefit alone, the case is interesting but not yet made. The epigenetic clock signal is the strongest pharmacological aging-biomarker result on record, but it's one trial, in one population, with one dose. Microdosing is plausible, regulatorily grey, and largely untested.

For an adult with overweight/obesity, established cardiovascular risk, or metabolic dysfunction, the case is much stronger: hard MACE reduction, durable weight loss, plus the longevity-relevant biomarker shifts on top.

In either case, the single most important sentence in this article: a GLP-1 without protein and resistance training is a slow path to sarcopenic obesity, not longevity. If you are not willing to lift weights and hit 1.6–2.2 g/kg protein/day, do not start. If you are, then it becomes one of the more interesting tools in current geroscience.

Further reading

• SELECT — semaglutide reduces MACE ~20% (HR 0.80) in about 17,600 non-diabetic adults with overweight/obesity and CVD; hard-outcome RCT, Novo Nordisk-funded.^[27]
• SELECT 4-year sustained weight and CV benefits.^[28]
• Non-diabetic semaglutide weight-loss meta-analysis (7 RCTs, n=5,411).^[29]
• OASIS 4 — oral semaglutide for non-diabetic weight loss.^[30]
• Wharton S et al. Gastrointestinal tolerability of once-weekly semaglutide 2.4 mg in adults with overweight or obesity. Diabetes Obes Metab 2022.^[31]
• He L et al. Association of GLP-1 Receptor Agonist Use With Risk of Gallbladder and Biliary Diseases: systematic review and meta-analysis of 76 RCTs. JAMA Intern Med 2022.^[32]
• Sodhi M et al. Risk of Gastrointestinal Adverse Events Associated With GLP-1 Receptor Agonists for Weight Loss. JAMA 2023.^[33]
• Hathaway JT et al. Risk of Nonarteritic Anterior Ischemic Optic Neuropathy in Patients Prescribed Semaglutide. JAMA Ophthalmol 2024.^[34]
• Hsu AY et al. Semaglutide and Nonarteritic Anterior Ischemic Optic Neuropathy Risk Among Patients With Diabetes. JAMA Ophthalmol 2025.^[35]
• Urbina S et al. Micronutrient and Nutritional Deficiencies Associated With GLP-1 Receptor Agonist Therapy: A Narrative Review. Clin Obes 2026.^[36]
• SURMOUNT-5 — tirzepatide vs. semaglutide head-to-head, 20.2% vs. 13.7% weight loss (~47% greater relative loss); RCT on a surrogate endpoint, Eli Lilly-funded.^[37]
• Semaglutide and validated epigenetic aging clocks — first RCT in HIV-lipohypertrophy.^[38]
• GLP-1 RAs across the twelve hallmarks of aging — translational review.^[39]
• STEP-HFpEF — exercise capacity in obesity-related HFpEF.^[40]
• STRIDE — semaglutide and walking capacity in symptomatic peripheral artery disease with type 2 diabetes; phase 3b RCT.^[41]
• EVOKE / EVOKE+ — primary endpoint missed in early Alzheimer's, biomarkers moved.^[42]
• Taipale H et al. Association between GLP-1 receptor agonist use and worsening mental illness in people with depression and anxiety in Sweden: a national cohort study. Lancet Psychiatry 2026.^[43]
• BELIEVE — semaglutide + bimagrumab, 92.8% of weight lost was fat; phase 2 RCT.^[44]
• Hansen MS et al. Once-weekly semaglutide versus placebo in adults with increased fracture risk: a phase 2 trial (hip BMD −2.6%, lumbar −2.1%, tibial cortical −1.8%). EClinicalMedicine 2024.^[45]
• Liu Y et al. Skeletal effect of semaglutide and tirzepatide — bone loss greater in non-diabetics than in type 2 diabetes. J Clin Endocrinol Metab 2026.^[46]
• Nutritional priorities advisory (ACLM/ASN/OMA/OS).^[47]
• LEAN-PREP muscle-preservation protocol trial.^[48]
• Wilding JPH et al. Weight regain after withdrawal of semaglutide: STEP 1 extension. Diabetes Obes Metab 2022.^[49]
• GLP-1 compounding and microdosing safety concerns.^[50]

---