insulin

Q: What is insulin human?

Insulin human is the body's own glucose-regulating hormone, produced for pharmaceutical use by recombinant DNA in a non-pathogenic laboratory strain of Escherichia coli. The Humulin R label gives the molecular formula as C257H383N65O77S6 and the molecular weight as 5808 Da. It is a licensed biological product, not a research peptide, with six FDA labels and two EMA authorisations covering diabetes mellitus.

Q: What does insulin do?

Section 12.1 of the labels states it directly: the primary activity of insulin is the regulation of glucose metabolism, lowering blood glucose by stimulating peripheral glucose uptake, especially by skeletal muscle and fat, and by inhibiting hepatic glucose production. The Novolin R label adds that insulin inhibits lipolysis and proteolysis and enhances protein synthesis, which is a mechanism statement inside a glycemic-control label rather than an outcome from a muscle trial.

Q: How is insulin human administered in the label record?

Subcutaneously for the regular human insulin products, intravenously under medical supervision for some presentations, by oral inhalation for Afrezza, and as a ready-to-use intravenous bag for Myxredlin. Every one of those routes exists inside a diabetes indication with glucose monitoring attached. There is no label route, schedule or dose for muscle or lean mass, because there is no such indication on any of these products.

Q: Is insulin human FDA approved?

Yes, for diabetes mellitus. Humulin R's original NDA 018780 was approved on 28 October 1982 and Novolin R under BLA 019938 in 1991; on 23 March 2020 every approved insulin NDA was deemed a biologics licence under the BPCI Act transition. On the EMA side, Insuman was authorised on 21 February 1997 and Actrapid on 7 October 2002. There is no approval, indication or efficacy claim anywhere in that record for muscle, hypertrophy or lean mass in people without diabetes.

Q: What are the side effects of insulin human?

The label names them. Section 5.3, hypoglycemia: severe hypoglycemia can cause seizures, may lead to unconsciousness, may be life threatening or cause death. Section 5.6, hypokalemia: all insulins cause a shift of potassium from the extracellular to the intracellular space, and untreated hypokalemia may cause respiratory paralysis, ventricular arrhythmia and death. The label also covers hypersensitivity reactions, fluid retention and heart failure with concomitant PPAR-gamma agonists, and dosing-error hypoglycemia. Lipohypertrophy at the site is common with repeated administration: in 500 patients injecting for at least two years it was found in 44.6% on examination with a further 13.4% picked up only on ultrasound. Injected regular human insulin carries no boxed warning; the inhaled product Afrezza does, and it is for bronchospasm in chronic lung disease.

Q: Why is insulin graded F here when it is called S for diabetes?

Because the two answer different questions and the page states both. This board grades the question that brings a reader to it, and that question is a research stack. In a stack insulin is an F: across six FDA labels and two EMA authorisations there is no muscle or lean-mass indication, two independent literature sweeps for a hypertrophy or lean-mass trial in healthy non-diabetic trained subjects returned nothing, and severe hypoglycemia is fast and documented. For diabetes, insulin is an S, one of the most important drugs in the history of medicine, on the WHO Model List of Essential Medicines, with a century of use behind it. The F is a warning about a specific misuse rather than a rating of the medicine.

two grades, and they are not close. S for diabetes, where it is one of the most important drugs in the history of medicine. F in a research stack, where the muscle record is empty and the harm is documented.

tier F · growth hormone · 0 muscle trials · S for diabetes

verdict

this entry carries two grades. for diabetes it is an S, and that is not a courtesy: six FDA labels, two EMA authorisations, a place on the WHO Model List of Essential Medicines, and more lives saved than anything else that will ever appear on this board. in a research stack it is an F, because across all six of those labels and both authorisations there is no indication, endpoint or efficacy claim for muscle, hypertrophy or lean mass in people without diabetes, and against that empty record sits severe hypoglycemia, which is fast and has no forgiving dose-finding window.

for anyone asking whether insulin belongs in a research stack — no. two independent literature sweeps looked for a controlled human trial of exogenous insulin with a hypertrophy or lean-mass endpoint in healthy non-diabetic trained subjects and returned nothing. the physique-specific human record is three single-patient case reports, one qualitative interview study of ten people, and one cross-sectional bloodwork comparison whose insulin subgroup is fifteen people, fourteen of them also on anabolic steroids and eleven of them also on growth hormone. against that, the label attaches the word death to two separate section headings. F on this board reads caution: evidence of harm, or evidence of nothing. insulin in a stack is both at once.

for anyone asking about insulin as medicine for diabetes — S, and the F on the card is not a comment on it. a century of use sits behind it. the modern record is six FDA labels including one whose original NDA was approved on 28 october 1982, two EMA authorisations dating to 1997 and 2002, a place on the WHO Model List of Essential Medicines, and trial programmes that still randomise a thousand people at a time: ONWARDS 1 put 984 adults with type 2 diabetes on weekly icodec against daily glargine and beat it on HbA1c by 0.19 points (95% CI -0.36 to -0.03, P=0.02). this page did not re-collect a century of diabetes evidence, because it does not need to. anyone managing diabetes should be following a prescriber, not a peptide board.

for anyone asking what insulin actually does to muscle protein — it is permissive and antiproteolytic rather than dose-dependently anabolic, and the human numbers are unusually clean on this. in seven healthy men under forearm perfusion at 124 microU/ml, uptake into protein did not move (phenylalanine 43 to 42, leucine 113 to 124) while release from protein fell hard (phenylalanine 57 to 33, leucine 126 to 63, P<0.01). raising clamped insulin from 5 to 30 mU/l halved leg protein breakdown; going on to 72 and then 167 mU/l bought nothing further, and muscle protein synthesis at all three of those concentrations was identical to the rate at 5. protein synthesis moves with amino acid availability, not with more units.

for anyone asking whether this is a vial-quality problem — it is not. regular human insulin is classified in the federal drug listing as a human OTC drug product. a survey of 582 US pharmacies, 561 responding, found Walmart locations selling it over the counter daily at 87.0% of stores, at $24.88 per 10 mL vial, with a median of four vials a day per store. the failure mode in the record is a correctly labelled pharmacy product with no glucose monitor attached to it.

based on FDA and EMA label records, published trials and the peer-reviewed literature. educational only, not medical advice. anyone managing diabetes should be following a prescriber.

why F-tier

F, and on this entry the letter works as a warning rather than a rating. the board grades the question that brings a reader to it, and that question is a research stack. F here reads caution: evidence of harm, or evidence of nothing. insulin in a stack is both at once. the evidence of nothing: across six FDA labels (Humulin R U-100, Humulin R U-500, Novolin R, Myxredlin, Afrezza, Awiqli) and two EMA authorisations (Actrapid, Insuman) there is no indication, endpoint or efficacy claim for muscle, hypertrophy or lean mass in people without diabetes; every indication on every one of them is diabetes mellitus. two independent literature sweeps looked for a controlled human trial with a hypertrophy or lean-mass endpoint in healthy non-diabetic trained subjects and returned nothing, and the registry sweep that would have confirmed that absence rather than inferred it was quota-blocked and could not run, so the claim here is an absence in the published literature and not a registry-confirmed one. what human mechanism data exists does not say dose-dependently anabolic: seven perfused forearms at 124 microU/ml showed net anabolism from suppressed breakdown with uptake into protein unchanged, and clamping insulin at 30, 72 and 167 mU/l produced the same muscle protein synthesis rate as clamping it at 5, with the effect on breakdown ceilinged at 30. insulin lifts muscle protein synthesis only where amino acid availability is already elevated, and the authors who tested the incremental effect consistently failed to find one. the evidence of harm: two label section headings carry the word death, hypoglycemia and hypokalemia, and 22 poisoned patients given high-dose insulin in critical care with pre-emptive glucose and electrolytes still went hypoglycemic 16 times out of 22 and hypokalemic 18 times out of 22. hypoglycemia is fast and there is no forgiving dose-finding window. D is wrong because D means hype outrunning data, and that framing needs a data side to outrun. and none of this is a grade on insulin as medicine: for diabetes it is an S, one of the most important drugs in the history of medicine, on the WHO Model List of Essential Medicines, with a century of use and a thousand-patient modern trial programme behind it.

the core tension

both grades are true at the same time. for diabetes the evidence is among the strongest in the history of drug development: a century of use, six FDA labels, two EMA authorisations, a place on the WHO Model List of Essential Medicines. and in the specific question this board grades, the record is empty. across all of that, not one indication or endpoint for muscle, hypertrophy or lean mass in people without diabetes. the single sentence the forums quote is a mechanism clause in section 12.1 of a glycemic-control label, and the human turnover work reads it narrowly: net anabolism in seven perfused forearms came from suppressed breakdown with synthesis unchanged, and clamping insulin at 30, 72 and 167 mU/l produced the same muscle protein synthesis rate as clamping it at 5. the harm does not wait for the argument to be settled: severe hypoglycemia is fast, and there is no forgiving dose-finding window to feel it out in.

what it is

Insulin human is the body's own glucose-regulating hormone, made for pharmaceutical use by recombinant DNA in a non-pathogenic laboratory strain of Escherichia coli. The Humulin R label gives the molecular formula as C257H383N65O77S6 and the molecular weight as 5808 Da, in vials of 100 units per mL. It is a licensed biological product rather than a research peptide, with six FDA labels, two EMA authorisations, and a first US approval dating to 1982.

what it does

Section 12.1 of the labels states the job in one sentence: "The primary activity of insulin, including HUMULIN R, is the regulation of glucose metabolism. Insulin lowers blood glucose by stimulating peripheral glucose uptake, especially by skeletal muscle and fat, and by inhibiting hepatic glucose production." The Novolin R label adds a further mechanism clause: "Insulin inhibits lipolysis and proteolysis, and enhances protein synthesis." That clause is a mechanism statement inside a glycemic-control label. It is not an outcome from a hypertrophy trial, and the human protein-turnover work reads it more narrowly than the forums do.

origin

The molecule predates almost everything else graded here, with a century of use behind it. The recombinant era began with Humulin R, whose original NDA 018780 was approved on 28 October 1982. Novolin R followed under BLA 019938 in 1991. On 23 March 2020 every approved insulin NDA was deemed a biologics licence under the BPCI Act transition, so Humulin R and Novolin R now sit in the federal record as BLA 018780 and BLA 019938. Concentrated U-500 regular insulin has been available in the US since 1952.

why researchers are interested

In physique circles insulin travels with growth hormone, and the reason it travels there is a single mechanism sentence in a diabetes label plus the fact that recombinant human insulin is sequence-identical to the body's own and therefore not distinguishable by current anti-doping methods. Prevalence estimates are consistent across two sources: a 500-person survey of anabolic steroid users reported 25% admitting adjuvant growth hormone and insulin, and a 2024 cross-sectional study of 92 recreational bodybuilders found insulin among 38% of the 40 who reported illicit hormone use.

does it work

For diabetes, comprehensively, and that is a separate grade. For muscle in people without diabetes, the controlled human record is absent rather than weak, which is a different and harsher statement. Two independent sweeps searched for a trial of exogenous insulin with a hypertrophy or lean-mass endpoint in healthy non-diabetic trained subjects and found none. The registry sweep that would have confirmed absence rather than inferred it was quota-blocked and could not run, so this is an absence in the published literature and not a registry-confirmed absence. What does exist is mechanism work, and it does not say dose-dependently anabolic. It says permissive and antiproteolytic, with a ceiling around 30 mU/l past which more insulin changed nothing in muscle protein turnover in healthy young men.

claims vs the data

insulin is anabolic, so more units build more muscle — contradicted — Greenhaff clamped insulin at 5, 30, 72 and 167 mU/l in healthy young men. Going from 5 to 30 halved leg protein breakdown (P<0.05); above 30 there was no further inhibition, and muscle protein synthesis at 30, 72 and 167 was identical to the rate at 5. Tripling amino acid availability at 5 mU/l doubled synthesis on its own. More insulin bought nothing.
the FDA label itself says insulin enhances protein synthesis — partially true — It does, in section 12.1 of the Novolin R label, as a mechanism statement inside a glycemic-control label: "Insulin inhibits lipolysis and proteolysis, and enhances protein synthesis." It is not an outcome from a hypertrophy trial. Every indication on every insulin label reviewed here is diabetes mellitus, and Gelfand's forearm work found uptake into protein unchanged while release from protein fell hard.
insulin is undetectable in drug testing — partially true — Recombinant human insulin is sequence-identical to endogenous insulin and current anti-doping methods cannot distinguish it, which is the accurate half. The analogs are detectable: nano-LC high resolution mass spectrometry gives a limit of detection of 10 pg/mL and a urinary window of roughly 9 hours for lispro, aspart and glulisine. The practical discriminator for human insulin is the insulin-to-C-peptide ratio, not the molecule, since exogenous insulin raises insulin without raising C-peptide.
insulin carries an FDA boxed warning for hypoglycemia — contradicted — It does not. Humulin R U-100, Humulin R U-500, Novolin R, Myxredlin and Awiqli all carry no boxed warning. The one insulin human product that does is Afrezza, the inhaled powder, and its box is for acute bronchospasm in chronic lung disease, with contraindications in asthma and COPD. That warning does not transfer to the subcutaneous products.
a hypoglycemic episode is manageable, it just needs sugar — contradicted — In 22 poisoned patients given high-dose insulin deliberately in a critical care setting, with glucose and electrolyte replacement running pre-emptively the whole time, 16 of 22 (73%) still developed hypoglycaemia, severe in 9, and 18 of 22 (82%) developed hypokalaemia. Glucose had to continue a median of 18 hours after the insulin was stopped. One case report needed glucose plus glucagon and intubation after a single 70 IU dose; another needed 14 days and a central line after 2100 units.
bodybuilders who research insulin are visibly bigger, so it works — weak — The one cross-sectional comparison that exists found fat-free mass index 26.2 kg/m2 in the insulin subgroup against 22.3 in non-hormone bodybuilders and 20.4 in controls (p<0.001). It cannot attribute that to insulin: the subgroup is 15 people, 14 of whom also reported anabolic steroids and 11 of whom also reported growth hormone, and the design is cross-sectional. The same study's authors note that trials testing whether exogenous insulin adds to muscle protein synthesis during hyperaminoacidaemia consistently failed to find an incremental effect.
insulin is a dangerous drug the site is warning people off — contradicted — For diabetes, insulin is an S: six FDA labels, two EMA authorisations, a place on the WHO Model List of Essential Medicines, a century of use, and ONWARDS 1 randomising 984 people on a single modern question. The F grades a research stack and nothing else.
the risk here is vial quality, like everything else on this board — overreach — Regular human insulin is classified in the federal drug listing as a human OTC drug product, and a 561-pharmacy survey found 87.0% of Walmart locations selling it over the counter daily at $24.88 per 10 mL vial. It is legal without a prescription in 49 states under pre-1938 grandfathering; Indiana banned the practice in 2014. The product arrives correctly labelled. The missing piece is the glucose monitor and the prescriber.

key facts

molecular formula: C₂₅₇H₃₈₃N₆₅O₇₇S₆
molecular weight: 5808 Da (label writes 5.808 kDa)
amino acids: not stated in the sourced label records; section 11 gives the molecular formula instead
half-life: 5 to 10 minutes in circulation; Evans 2003 gives about 4 minutes. subcutaneous action outlasts the molecule
type: human insulin of recombinant DNA origin (rDNA)
CAS: not in the sourced records (FDA UNII 1Y17CTI5SR; RxCUI 311033, 311034, 311036)
6 FDA labels, zero muscle indications
0 lean-mass trials in healthy adults
30 mU/l where the effect stops climbing
0.6 mmol/l glucose in the published case

frequently asked questions

What is insulin human?

Insulin human is the body's own glucose-regulating hormone, produced for pharmaceutical use by recombinant DNA in a non-pathogenic laboratory strain of Escherichia coli. The Humulin R label gives the molecular formula as C257H383N65O77S6 and the molecular weight as 5808 Da. It is a licensed biological product, not a research peptide, with six FDA labels and two EMA authorisations covering diabetes mellitus.

What does insulin do?

Section 12.1 of the labels states it directly: the primary activity of insulin is the regulation of glucose metabolism, lowering blood glucose by stimulating peripheral glucose uptake, especially by skeletal muscle and fat, and by inhibiting hepatic glucose production. The Novolin R label adds that insulin inhibits lipolysis and proteolysis and enhances protein synthesis, which is a mechanism statement inside a glycemic-control label rather than an outcome from a muscle trial.

How is insulin human administered in the label record?

Subcutaneously for the regular human insulin products, intravenously under medical supervision for some presentations, by oral inhalation for Afrezza, and as a ready-to-use intravenous bag for Myxredlin. Every one of those routes exists inside a diabetes indication with glucose monitoring attached. There is no label route, schedule or dose for muscle or lean mass, because there is no such indication on any of these products.

Is insulin human FDA approved?

Yes, for diabetes mellitus. Humulin R's original NDA 018780 was approved on 28 October 1982 and Novolin R under BLA 019938 in 1991; on 23 March 2020 every approved insulin NDA was deemed a biologics licence under the BPCI Act transition. On the EMA side, Insuman was authorised on 21 February 1997 and Actrapid on 7 October 2002. There is no approval, indication or efficacy claim anywhere in that record for muscle, hypertrophy or lean mass in people without diabetes.

What are the side effects of insulin human?

The label names them. Section 5.3, hypoglycemia: severe hypoglycemia can cause seizures, may lead to unconsciousness, may be life threatening or cause death. Section 5.6, hypokalemia: all insulins cause a shift of potassium from the extracellular to the intracellular space, and untreated hypokalemia may cause respiratory paralysis, ventricular arrhythmia and death. The label also covers hypersensitivity reactions, fluid retention and heart failure with concomitant PPAR-gamma agonists, and dosing-error hypoglycemia. Lipohypertrophy at the site is common with repeated administration: in 500 patients injecting for at least two years it was found in 44.6% on examination with a further 13.4% picked up only on ultrasound. Injected regular human insulin carries no boxed warning; the inhaled product Afrezza does, and it is for bronchospasm in chronic lung disease.

Why is insulin graded F here when it is called S for diabetes?

Because the two answer different questions and the page states both. This board grades the question that brings a reader to it, and that question is a research stack. In a stack insulin is an F: across six FDA labels and two EMA authorisations there is no muscle or lean-mass indication, two independent literature sweeps for a hypertrophy or lean-mass trial in healthy non-diabetic trained subjects returned nothing, and severe hypoglycemia is fast and documented. For diabetes, insulin is an S, one of the most important drugs in the history of medicine, on the WHO Model List of Essential Medicines, with a century of use behind it. The F is a warning about a specific misuse rather than a rating of the medicine.

related peptides

igf-1 lr3 — the other F, and the other compound whose hypoglycemia risk comes from the insulin receptor
hgh — the compound insulin travels with in these circles, and the one the 70 IU case was chasing
mots-c — same WADA class, different subsection: S4.4.1 rather than S4.4.2
semaglutide — the other end of glucose control, and an S on a question this board does grade

reptides grades the research record and cites the literature behind every call. research reference only; not medical advice.