Harm mitigation

Retatrutide: the escalation is the safety protocol

Retatrutide is the compound in this catalogue where the gap between what was studied and what is commonly done is widest, so it gets the most space here. It is a triple agonist — GLP-1, GIP and glucagon — and the glucagon arm is what separates it from everything before it. That arm raises energy expenditure, and it also raises heart rate more than a GLP-1-only agonist does. Both effects scale with dose.

The phase 2 obesity trial escalated monthly: 2 mg, 4 mg, 8 mg, 12 mg. The phase 3 programme uses 2, 4, 6, 9, 12. Those are the two ladders that exist. A six-step ladder running 2/4/6/8/10/12 circulates widely and is sometimes attributed to the NEJM paper; it is not in that paper, and a 10 mg maintenance dose appears in no retatrutide trial at all. If a schedule cannot be traced to a published trial, the adverse-event rates people quote alongside it cannot be either.

The reason this matters more than it would for a gentler compound is that retatrutide's tolerability is escalation-rate-dependent, not simply dose-dependent. The same 8 mg reached over four months and reached over four weeks are not equivalent exposures. Nausea, vomiting and diarrhoea in the trials clustered around dose increases and settled at a stable dose — which is precisely why the schedule holds each step for four weeks before moving.

The practical consequence: a step that produces significant GI effects is a step to stay on, not to push through. Going up again while still symptomatic stacks a new escalation on an unresolved one, and the resulting dehydration is the mechanism behind most of the genuinely dangerous outcomes in this class — acute kidney injury, electrolyte disturbance, and the syncope that follows both.

• Hold a dose for at least four weeks before increasing, and longer if the last increase is still producing symptoms.
• Never make up a missed week by doubling. The half-life is around six days, so a missed dose is a gap, not a debt.
• Heart rate rises as a matter of mechanism, not as a side effect to be surprised by. A persistent resting rise of 10+ bpm is worth taking seriously rather than normalising.
• Vomiting or diarrhoea that prevents normal fluid intake is a stop-and-rehydrate situation, not a taper-later one.

What the incretin class does to the rest of the body

These effects apply across retatrutide, tirzepatide, semaglutide and — for the amylin arm — cagrilintide. They are class-level and none of them have been excluded for the newer compounds; the absence of a signal in a short trial is not the same as the absence of the effect.

Delayed gastric emptying is the mechanism behind the satiety and also behind the risk. It changes the absorption timing of anything taken by mouth, and it is the reason anaesthetic guidance now asks about GLP-1 use before sedation — retained stomach contents after a normal fast is an aspiration risk. This is the single interaction most likely to matter and the one least likely to be volunteered.

Gallbladder disease rises with rapid weight loss regardless of cause, and the incretins produce rapid weight loss. Pancreatitis is a class warning that remains unresolved rather than excluded. Both present as upper abdominal pain that is different in character from the ordinary nausea of a dose increase — severe, persistent, often radiating to the back. That distinction is the one worth being able to make.

Muscle loss is the quiet one. A substantial fraction of the weight lost on these compounds is lean mass, and nothing about the mechanism protects against it. Protein intake and resistance training are the only things shown to blunt it, and neither is optional at the higher doses.

• Tell an anaesthetist or endoscopist before any procedure requiring sedation, even if the last dose was a week ago.
• Severe persistent upper abdominal pain, particularly radiating to the back, is a stop-and-seek-care sign, not a dose-adjustment one.
• Combining with insulin or a sulfonylurea introduces real hypoglycaemia risk. The incretins alone rarely cause it; stacked with a secretagogue they do.
• Do not stack two incretins for additive effect. They share receptors, and the adverse effects add faster than the benefit.

Growth hormone secretagogues

CJC-1295, ipamorelin, tesamorelin and the rest of the secretagogue family work through the pituitary, which is their principal safety advantage: feedback limits the achievable GH level in a way that exogenous hGH does not. The ceiling is real, but it is a ceiling on peak GH, not on the consequences of sustained elevation.

The reliable effects are glucose-side. GH is counter-regulatory to insulin, so sustained stimulation raises fasting glucose and can push an already-marginal insulin sensitivity into a worse place. Water retention, carpal tunnel symptoms and joint aches are the common dose-limiting complaints, and all three resolve on dose reduction.

The unresolved question is proliferative. GH and IGF-1 are growth signals, and there is no trial establishing what years of intermittent elevation does to a pre-existing malignancy or a predisposition to one. This is the reason cycling — commonly 8–12 weeks on and 4–6 weeks off — exists in reported practice, rather than the receptor-desensitisation argument the evidence does not really support.

• Fasting glucose is the marker that actually moves. Checking it before starting gives the later number something to mean.
• Carpal tunnel symptoms and persistent oedema are dose signals — reduce rather than push through.
• An active or historical malignancy is the clearest reason not to use this class at all.
• Exogenous hGH removes the pituitary feedback ceiling entirely and belongs in a different risk category from the secretagogues.

GHK-Cu and the copper ceiling

GHK-Cu is the one compound here where the limit comes from something other than peptide pharmacology. Each molecule carries a copper ion, the complex is roughly 16% copper by mass, and copper accumulates. Fifty milligrams of GHK-Cu carries in the region of 8 mg of copper against a typical dietary intake of 1–2 mg a day.

This is why the reported protocols are short blocks with real breaks — 2–4 weeks on, 2–4 weeks off — and why total monthly exposure matters more here than the daily figure. The same logic governs the GLOW and KLOW blends, where GHK-Cu is the component setting the ceiling for the whole blend.

The visible sign of overexposure is a bluish-grey discolouration at repeatedly used injection sites. Systemic copper excess is a more serious matter and presents non-specifically — fatigue, nausea, abdominal pain — which is exactly why the exposure limit is worth respecting prospectively rather than reacting to.

Repair peptides and the angiogenesis question

BPC-157 and TB-500 have an unusually clean short-term profile in the animal literature and essentially no long-term human data. The honest position is that their principal risk is unknown rather than low.

The specific concern is mechanistic. Both promote angiogenesis, and a compound that helps a tendon build blood supply is not selective about what else might want one. There is no evidence of harm in humans; there is also no study designed to find it. Anyone with an active malignancy or an undiagnosed mass is taking an unquantified risk that the tendon benefit does not offset.

Sterility, technique and the risks you introduce yourself

The compound is only half of the risk. The other half is introduced during reconstitution and injection, and unlike the pharmacology it is almost entirely controllable.

Bacteriostatic water contains 0.9% benzyl alcohol, which suppresses bacterial growth for the 28-day window; sterile water does not, and a vial reconstituted with it should be treated as single-session. Neither makes a contaminated vial safe. The realistic failure is a needle that touched a surface, a stopper that was not wiped, or a vial kept past the window because there was material left in it.

Injection-site abscesses are the most common real-world harm across every compound in this catalogue, and they come almost entirely from technique rather than chemistry. Site rotation matters for the same reason: repeatedly injecting one area produces lipohypertrophy, which then changes absorption unpredictably and makes the dose itself less reliable.

• Wipe the stopper every time, not only on first puncture.
• One needle, one use. Reused needles are blunt, and blunt needles cause the tissue trauma that abscesses grow in.
• Rotate sites. Absorption from lipohypertrophic tissue is erratic, which turns a known dose into an unknown one.
• Respect the 28-day reconstituted window and reconstitute only the portion a block will actually use.
• Spreading warmth, redness and pain at a site is an infection until proven otherwise — it needs care, not another injection elsewhere.

Knowing what is in the vial

Every risk discussed above assumes the vial contains what the label says. Where it does not, none of the reasoning holds — a dose calculated for a compound that is not present, or is present at a different concentration, is not a dose at all.

A certificate of analysis for the specific batch, tied to the vial in front of you rather than to the product in general, is the only practical check available outside a laboratory. Purity is the headline number, but identity is the one that matters more: a 99% pure something-else is worse than a 95% pure correct compound.

When to stop

Most adverse effects in this catalogue are dose-related and resolve on reduction. A small number do not, and the distinction is worth knowing in advance rather than deciding in the moment.

• Severe persistent abdominal pain, especially radiating to the back — stop and seek care.
• Signs of an allergic reaction: facial or throat swelling, widespread rash, difficulty breathing.
• Vomiting or diarrhoea severe enough to prevent fluid intake.
• Chest pain, or a sustained resting heart rate well above baseline.
• Fever with a hot, spreading, painful injection site.
• Visual disturbance, severe headache, or fainting.
• Yellowing of the skin or eyes.