Peptide Health Lab

Metabolic & GLP-1

L-Carnitine

Also known as: Levocarnitine, Carnitor

Regulatory status
FDA approved
Evidence grade
Human RCT evidence

Last reviewed September 1, 2026 · 20 sources

What L-Carnitine is and how it works

L-carnitine is a small quaternary-ammonium compound that the body both synthesizes and takes in from food, and its job is transport. Long-chain fatty acids cannot cross the inner mitochondrial membrane on their own; carnitine carries them across as acylcarnitine esters so they can be broken down for energy. Cells accumulate it and the kidney holds on to it through OCTN2, a high-affinity organic cation transporter specific for carnitine, and a defect in the gene encoding that transporter produces autosomal recessive primary carnitine deficiency. In the diet it comes chiefly from meat and dairy products.[1]

Two pharmacokinetic facts do most of the work in reading everything below. First, after oral doses of 1 to 6 g the absolute bioavailability of L-carnitine is 5 to 18%, against a figure that may be as high as 75% for the smaller amounts present in a normal diet — a supplemental dose is absorbed markedly less efficiently than a dietary one. Second, under baseline conditions roughly 98 to 99% of filtered carnitine is reabsorbed by the renal tubule, so the plasma pool is defended tightly and renal clearance rises toward the glomerular filtration rate once exogenous carnitine pushes past a threshold around 40 to 60 µmol/L.[2] One consequence is that at pharmacological doses carnitine enters cells through the amino-acid transporter B(0,+) rather than through OCTN2 — the transport route that carries a supplemental dose is not the one that carries a dietary one.[1]

Three naming distinctions decide whether a given study is about this compound at all. Acetyl-L-carnitine and propionyl-L-carnitine are different molecules with their own separate literatures, and results from either do not transfer here. D-carnitine is not L-carnitine, and the approved labeling notes that mild myasthenia has been described only in uremic patients receiving the racemic D,L- form. And L-carnitine-L-tartrate, the salt most exercise trials administered, is a different mass from an equal weight of free L-carnitine — a dose quoted from a tartrate trial has to say tartrate. Beyond the molecule, the route and the population decide as much again: intravenous levocarnitine given to a patient on dialysis whose carnitine is being lost through the dialyser is not oral supplementation in an adult whose carnitine pool is already full.

What the research actually shows

The approved indication is repletion of a pool that is genuinely empty. A structured review assembled 757 individual published patients with primary carnitine deficiency from 166 articles. Among the 621 whose diagnosis rested on genetic or functional testing, cardiac involvement — predominantly cardiomyopathy — was the most prevalent presentation at 23.8%, with metabolic, hepatic, and neurological presentations occurring mainly in early childhood. The authors also record that almost 20% of the whole case literature, 136 patients, was diagnosed on a low carnitine concentration alone, which they treat as an uncertain diagnosis.[3]

In dialysis, the randomized evidence is large and mostly null. A meta-analysis of 49 randomized trials in 1,734 adults on maintenance hemodialysis found significant reductions in LDL cholesterol and C-reactive protein and no significant difference in seven other endpoints, including hemoglobin and the required erythropoietin dose — results the authors say failed to confirm an earlier synthesis. They call the LDL reduction not clinically relevant and the C-reactive protein reduction both statistically and clinically relevant, while noting that its relation to hard endpoints such as mortality remains to be clarified.[4]

Weight loss is real, randomized, and about one kilogram. A meta-analysis of 43 randomized trials reported a weighted mean difference of −1.129 kg in body weight, −0.359 kg/m² in BMI, and −1.158 kg in fat mass, with body fat percentage and waist circumference both non-significant, the effect confined to overweight and obese subjects, and anthropometric indexes unchanged among patients undergoing hemodialysis.[5] An earlier meta-analysis of nine trials in 911 adults reported −1.33 kg and found on meta-regression that the magnitude of weight loss significantly decreased with the duration of consumption (p = 0.002).[6] Those two are not independent replications of each other: they are separate syntheses of an overlapping pool of the same published trials, with search windows that overlap almost entirely, so the second does not corroborate the first the way two new trials would.

Muscle carnitine can be raised, under conditions that are rarely reproduced. In a randomized, double-blind trial, 14 healthy men took L-carnitine-L-tartrate alongside a large carbohydrate load twice daily for 24 weeks while a control arm took the carbohydrate alone. Muscle total carnitine rose 21%; at 50% of maximal oxygen uptake the carnitine group used 55% less muscle glycogen; at 80% muscle lactate was 44% lower and pyruvate dehydrogenase activation 38% higher; and work output in a 30-minute performance trial rose 11% from baseline while the control arm did not change.[7] A second randomized double-blind trial in 14 men averaging 69 years, using the salt inside an insulinogenic beverage for 25 weeks alongside twice-weekly cycling, raised muscle carnitine 20% and whole-body fat oxidation 20% during moderate-intensity exercise, with no change in insulin-stimulated glucose disposal.[8] Both required about half a year and a substantial daily insulin stimulus. A single 3 g dose 90 minutes before a high-intensity workout in 20 trained men did nothing at all — 202.4 against 204.5 repetitions, p = 0.810, with no difference in perceived exertion or blood pressure.[9]

The cardiac literature is old and its headline result is marginal. A meta-analysis of 13 controlled trials in 3,629 patients after acute myocardial infarction reported a 27% reduction in all-cause mortality (odds ratio 0.73, 95% CI 0.54–0.99, P = .05), a 65% reduction in ventricular arrhythmias, and a 40% reduction in the development of angina, with no reduction in heart failure or reinfarction. The authors' own conclusion asks for large randomized trials in the modern era.[10]

Sperm motility is the one fertility endpoint with a positive pooled result. A meta-analysis of randomized trials found carnitines to be the only intervention that increased progressive motility, by 7.45%, alongside increases in total motility and morphology — reported for a pooled "carnitines" category rather than for L-carnitine specifically, on semen parameters rather than pregnancy, and with the authors cautioning on sample size and heterogeneity.[17] The dedicated randomized trial in this area gave 21 men 2,000 mg of L-carnitine together with 1,000 mg of L-acetyl-carnitine daily for 24 weeks and found no statistically or clinically significant increase in motility or total motile counts.[18]

Where the evidence is weak

The approval is for something almost nobody taking this is treating. Everything in the deficiency literature is repletion of a pool that is measurably low, and a result obtained in a depleted population does not transfer to a replete one. The weight meta-analysis makes the point from the other direction: the anthropometric effect it found in overweight and obese subjects was absent in the hemodialysis subgroup.[5]

A kilogram is a kilogram. The randomized weight effect is consistent and small, body fat percentage and waist circumference did not move, and the effect shrank as trials ran longer.[6] Counting the repeated meta-analyses of this literature as repeated confirmations overstates it; they are re-analyses of largely the same trials.

The loading protocol, not the carnitine alone, is what moved muscle carnitine. Both trials that raised muscle content paired the compound with a large daily insulin stimulus for roughly half a year, and neither tested the compound on its own.[7] Acute pre-exercise dosing, which is how it is most often taken, produced a clean null.[9]

Claims that belong to other molecules keep getting read across. The best-evidenced walking-distance result in this family is propionyl-L-carnitine's — moderate-certainty Cochrane evidence of about a 51 metre improvement in maximal walking distance in intermittent claudication, in trials the reviewers note were largely drug-company-backed — and Cochrane's only direct comparison against L-carnitine is a single 14-participant crossover at low certainty. There is no L-carnitine equivalent of that evidence base.[11]

Cognition is unstudied rather than disproved. A Cochrane review searching L-carnitine, acetyl-L-carnitine, and propionyl-L-carnitine terms found only two eligible randomized trials in cognitively healthy people, rated every outcome very low quality, and stated that it could draw no conclusion about efficacy or safety.[20]

Nothing has tested whether taking it changes cardiovascular events. PubMed searches on 1 September 2026 for L-carnitine with major adverse cardiovascular events or clinical endpoint restricted to randomized controlled trials returned no records, and a search for levocarnitine with mortality, myocardial infarction, or stroke restricted to randomized controlled trials returned two — a sepsis trial and a pulse-wave-velocity trial in hemodialysis patients, neither of which tested events in a replete adult. The exact query strings are recorded in this story's development log so the search can be re-run.

The metabolite signal is genuine, and so is the case against reading it as causal. Gut bacteria metabolize dietary L-carnitine to trimethylamine, which is oxidized to trimethylamine-N-oxide. In the landmark work, mice fed carnitine developed accelerated atherosclerosis and mice whose intestinal microbiota was concurrently suppressed did not, and in an observational cohort of 2,595 people undergoing cardiac evaluation reported in the same paper, plasma carnitine predicted prevalent cardiovascular disease and incident major adverse cardiac events — but only among subjects with concurrently high trimethylamine-N-oxide.[12] Supplementation reliably raises the metabolite: a 24-week randomized, double-blind, placebo-controlled trial in healthy women aged 65 to 70 reported roughly a tenfold rise, and in the same trial C-reactive protein, interleukin-6, tumour necrosis factor-α, L-selectin, P-selectin, both adhesion molecules measured, and the lipid profile were all unchanged.[16] Observational syntheses associate the metabolite with all-cause mortality at a hazard ratio of 1.91 across 17 studies and 26,167 subjects, with 94% heterogeneity on that outcome.[13] Against that, a bidirectional Mendelian randomization analysis found that genetically predicted trimethylamine-N-oxide and carnitine were not associated with type 2 diabetes, atrial fibrillation, coronary artery disease, myocardial infarction, stroke, or chronic kidney disease after correction, while type 2 diabetes and chronic kidney disease were causally associated with higher metabolite levels — the authors conclude the observational evidence may reflect confounding or reverse causality.[14] A 2025 umbrella review of 27 systematic reviews found the causal question controversial, prospective studies in healthy populations entirely absent, and serious methodological flaws leaving critically low confidence in the published syntheses.[15] The metabolite is cleared renally, which is exactly the direction the Mendelian analysis found: kidney disease raises it. Carrying only one half of this is not honest reporting in either direction.

Legal and regulatory status

Under the name levocarnitine, L-carnitine is an FDA-approved prescription drug — and the scope of that approval is the single most misread fact about this compound. Fourteen applications currently sit on the active ingredient. Three are innovator new drug applications held by Leadiant Biosciences under the CARNITOR brand — NDA 018948, approved 27 December 1985, whose 330 mg oral tablet is currently marketed as a prescription product while its oral solution is discontinued; NDA 019257, approved 10 April 1986, covering an oral solution and a sugar-free oral solution; and NDA 020182, approved 16 December 1992, covering a 200 mg/mL injectable. The other eleven are prescription abbreviated applications across the same three dosage forms, two of which carry discontinued products. A discontinued row is not a current approval.

The approved indications are the whole point. The oral labeling is indicated for the treatment of primary systemic carnitine deficiency and for the acute and chronic treatment of patients with an inborn error of metabolism resulting in secondary carnitine deficiency. The injectable labeling adds the prevention and treatment of carnitine deficiency in end-stage renal disease patients undergoing dialysis. That is the entire approved scope. There is no approved indication for weight loss, for athletic performance, for fertility, or for cognition, and an approval for carnitine deficiency cannot be read as an approval for any of them.

No boxed warning appears on any of the twenty-six labeling records returned for the ingredient, checked field by field against a control substance whose labels do carry one. The labeling does carry a hypersensitivity warning that includes anaphylaxis, laryngeal edema, and bronchospasm reported after intravenous administration, mostly in dialysis patients; a note that seizures have been reported in patients with or without pre-existing seizure activity receiving either oral or intravenous levocarnitine, with increased frequency or severity in those who already have it; a dose-related body-odor note; a precaution to monitor INR in patients on warfarin; the instruction that the oral solutions are for oral use only and not for parenteral use; and — the sentence most relevant to the section above — an explicit precaution that chronic administration of high oral doses in patients with severely compromised renal function or on dialysis may result in accumulation of the potentially toxic metabolites trimethylamine and trimethylamine-N-oxide, since these are normally excreted in the urine. Contraindications are listed as none known.

The split between an oral and a parenteral approved product is a pharmacological fact before it is a regulatory one. An intravenous dose distributes into an initial volume of roughly 0.2 to 0.3 L/kg, corresponding to the extracellular fluid volume, and drives renal clearance up toward the glomerular filtration rate, while an oral dose first has to clear an absorption step that a dietary amount largely does not.[2] The two routes are not interchangeable, and a dose given intravenously to a patient on dialysis does not read across to an oral one in anybody else.

The second half of this compound's regulatory position is what makes it unusual in this library: the same molecule is also sold over the counter as a dietary supplement. That is a statutory category, not a second approval — a dietary supplement is not reviewed or approved by the FDA before it goes on sale and is not required to demonstrate an effect. The two statuses coexist without informing each other. The prescription products appear in the national drug code directory alongside bulk active-ingredient registrations, which are registrations of a raw material rather than approvals of anything; dietary supplements do not appear in that directory at all, which is the cleanest demonstration of where the line falls. Peptide Health Lab does not sell anything and does not tell anyone where to obtain anything. In tested sport, the 2026 World Anti-Doping Code International Standard Prohibited List — in force from 1 January 2026, retrieved from the World Anti-Doping Agency on 1 September 2026 via the United States Anti-Doping Agency's prohibited-list page — does not name carnitine, L-carnitine or levocarnitine anywhere in its text, and the accompanying 2026 Monitoring Program, retrieved the same day, does not name it either. The provision that does bear on this compound attaches to a route rather than to a substance: prohibited method M2.2 prohibits intravenous infusions or injections of more than a total of 100 mL per 12-hour period at all times, except where legitimately received during hospital treatment, surgery, or clinical diagnostic investigation. That provision is about volume and route, not about this substance, and it applies whatever is in the bag. The list is revised annually, so anyone competing under anti-doping rules should read the year in force at first hand; a static page is not a reliable substitute for either document.

Questions to bring to a provider

Because a real approval and a real randomized literature both exist here, the useful questions are about which one applies:

  • Is there any reason to think a carnitine deficiency is present — a measured plasma free carnitine, a genetic result, an inborn error, or dialysis — because the approved indication and most of the strongest evidence are about repletion of a depleted pool?[3]
  • If the goal is weight, is a randomized average around one kilogram, with body fat percentage and waist circumference unchanged, worth pursuing this way?[5]
  • If the goal is performance, does the plan resemble the trials that worked — roughly six months alongside a large daily carbohydrate load — or the acute pre-workout dose that did nothing?[7][9]
  • How does existing kidney disease bear on this, given that the approved labeling warns about accumulation of trimethylamine and trimethylamine-N-oxide in severely compromised renal function, and that Mendelian analysis suggests kidney disease itself raises that metabolite?[14]
  • Is warfarin or any other anticoagulant in the picture, given that the approved labeling asks for INR monitoring after starting levocarnitine or changing the amount?
  • What weight should a history of seizures carry here, given that the approved labeling reports them in patients both with and without pre-existing seizure activity?
  • Given that oral absorption at supplemental doses runs at 5 to 18% and the kidney reabsorbs almost all of what is filtered, what would actually be measured to know whether anything changed?[2]
  • If the source of interest was a result in acetyl-L-carnitine or propionyl-L-carnitine, does that evidence say anything about this molecule at all?[11][20]

A clinician who says "the approval is real, and it is an approval for something you do not have" is reading this correctly.

Evidence by claim

Grades describe how strong the evidence is; the line under each grade describes what kind of studies it is. How we grade evidence.

Repletion in primary carnitine deficiency, the approved indication
Anecdotal reports only review

This is what the approval is for, and the published evidence behind it is case-based rather than randomized, which is ordinary for an ultra-rare inborn error. A structured review of 166 articles assembled 757 individual published patients; among the 621 whose diagnosis rested on genetic or functional testing, cardiac involvement was the most prevalent presentation at 23.8%, with metabolic, hepatic, and neurological presentations mostly in early childhood. Almost 20% of the total case literature — 136 patients — was diagnosed on a low carnitine concentration alone, which the authors treat as an uncertain diagnosis.

[3] [1]

Intravenous or oral levocarnitine in end-stage kidney disease on dialysis
Human RCT evidence review

Randomized evidence exists in quantity and is mixed. A meta-analysis of 49 randomized trials in 1,734 adults on maintenance hemodialysis found significant reductions in LDL cholesterol (−5.82 mg/dL) and C-reactive protein (−3.65 mg/L), and no significant difference in triglycerides, total cholesterol, HDL, hemoglobin, hematocrit, albumin, or the required erythropoietin dose. The authors state that they failed to confirm an earlier synthesis on hemoglobin and erythropoietin, call the LDL reduction not clinically relevant, and call the C-reactive protein reduction both statistically and clinically relevant while noting its relation to hard endpoints remains unclarified. This population loses carnitine through the dialyser, so these are repletion results.

[4] [2]

Weight loss in overweight and obese adults
Human RCT evidence review

Real, randomized, and small. A meta-analysis of 43 randomized trials reported a weighted mean difference of −1.129 kg in body weight (95% CI −1.590 to −0.669), −0.359 kg/m² in BMI, and −1.158 kg in fat mass, with no significant change in body fat percentage or waist circumference, an effect confined to overweight and obese subjects, and no change in anthropometric indexes among patients undergoing hemodialysis. An earlier meta-analysis of nine trials in 911 adults reported −1.33 kg and found in meta-regression that the magnitude of weight loss significantly decreased with the duration of consumption (p = 0.002). The two draw on an overlapping pool of the same published trials and are not independent replications of each other.

[5] [6]

Muscle carnitine loading and exercise performance
Human RCT evidence human RCT

Demonstrated, under conditions almost nobody reproduces. In 14 healthy men given L-carnitine-L-tartrate with a large carbohydrate load twice daily for 24 weeks, muscle total carnitine rose 21%, muscle glycogen use at 50% of maximal oxygen uptake fell 55%, muscle lactate at 80% fell 44%, and work output in a 30-minute performance trial rose 11% from baseline while the control arm did not change. In 14 men averaging 69 years, 25 weeks of the salt inside an insulinogenic beverage raised muscle carnitine 20% and whole-body fat oxidation 20% during moderate-intensity exercise, with no change in insulin-stimulated glucose disposal. A single 3 g acute dose 90 minutes before a high-intensity workout in 20 trained men changed nothing (202.4 versus 204.5 repetitions, p = 0.810).

[7] [8] [9]

Mortality and arrhythmia after acute myocardial infarction
Human RCT evidence review

A meta-analysis of 13 controlled trials in 3,629 patients reported a 27% reduction in all-cause mortality (odds ratio 0.73, 95% CI 0.54–0.99, P = .05), a 65% reduction in ventricular arrhythmias (risk ratio 0.35, P < .0001), and a 40% reduction in the development of angina (risk ratio 0.60, P < .00001), with no reduction in heart failure (risk ratio 0.85, P = .21) or reinfarction (risk ratio 0.78, P = .45). The mortality result sits exactly at the conventional threshold, the authors themselves call for large randomized trials in the modern era, and the finding is in patients after a myocardial infarction rather than in anyone taking a supplement.

[10]

Sperm motility in male infertility
Human RCT evidence review · human RCT

Positive on a surrogate, negative in the one trial that isolated it least badly. A meta-analysis of randomized trials found carnitines to be the only intervention that increased sperm progressive motility (7.45%, 95% CI 6.24–8.67), alongside increases in total motility and morphology — but the pooled category is carnitines rather than L-carnitine specifically, the endpoints are semen parameters rather than pregnancy or live birth, and the authors caution on limited sample sizes and considerable between-study heterogeneity. A separate randomized, double-blind, placebo-controlled trial in 21 men gave 2,000 mg of L-carnitine together with 1,000 mg of L-acetyl-carnitine for 24 weeks and found no statistically or clinically significant increase in motility or total motile counts — a two-molecule arm, which is another reason it is not a clean test of L-carnitine.

[17] [18]

Cognitive enhancement in people without cognitive impairment
Anecdotal reports only review

A Cochrane review searching L-carnitine, acetyl-L-carnitine, and propionyl-L-carnitine terms found only two eligible randomized trials, both in young adults, both poorly reported, and both at unclear risk of bias. No effect was seen on reaction time, vigilance, immediate memory, or delayed recall after three days of treatment in the one trial from which data could be extracted. The reviewers rated all outcomes very low quality and stated that they were unable to draw any conclusion about efficacy or safety. That is an absence of usable evidence rather than a demonstrated absence of effect.

[20]

Walking distance in intermittent claudication
Anecdotal reports only review

This claim belongs to a different molecule. The Cochrane review of intermittent claudication evaluates propionyl-L-carnitine, and found moderate-certainty evidence across 9 trials in 1,121 participants of a 50.86 metre improvement in maximal walking distance against placebo, noting a high proportion of drug-company-backed studies. Its only direct comparison of propionyl-L-carnitine against L-carnitine is a single 14-participant crossover at low certainty. There is no L-carnitine equivalent of that evidence base.

[11]

Whether supplementation raises trimethylamine-N-oxide
Human RCT evidence animal · human RCT

It does. Gut bacteria metabolize dietary L-carnitine to trimethylamine, which is oxidized to trimethylamine-N-oxide, and a 24-week randomized, double-blind, placebo-controlled trial in healthy women aged 65 to 70 reported roughly a tenfold rise in plasma trimethylamine-N-oxide. In the same trial, C-reactive protein, interleukin-6, tumour necrosis factor-α, L-selectin, P-selectin, and both adhesion molecules measured were unchanged, as was the lipid profile.

[12] [16]

Whether that metabolite rise causes cardiovascular harm
Anecdotal reports only animal · review · human observational · human RCT

Unestablished, and the evidence points in more than one direction. The landmark work showed accelerated atherosclerosis in mice fed carnitine and not in mice whose microbiota was suppressed, and reported in an observational cohort of 2,595 people undergoing cardiac evaluation that plasma carnitine predicted cardiovascular events only among those with concurrently high trimethylamine-N-oxide. Observational syntheses associate the metabolite with mortality at a hazard ratio of 1.91, with 94% heterogeneity. A bidirectional Mendelian randomization analysis found no causal association in the metabolite-to-disease direction and a significant one in reverse, concluding the observational signal may reflect confounding or reverse causality; an umbrella review of 27 syntheses reached the same unsettled place and found prospective studies in healthy populations entirely absent. Searches of the randomized-trial literature on 1 September 2026, recorded in the body of this page, surfaced no trial using a cardiovascular event as an endpoint.

[12] [13] [14] [15] [16]

Typical protocol range in the research

  • Randomized, double-blind trial in 14 healthy men over 24 weeks, in which the carnitine arm received the salt L-carnitine-L-tartrate together with a large carbohydrate load twice daily and the control arm received the carbohydrate alone

    2 g of L-carnitine-L-tartrate plus 80 g of carbohydrate, twice daily [7]

  • Randomized, double-blind trial in 14 healthy men aged 69 on average, over 25 weeks, with the daily dose given inside an insulinogenic beverage alongside twice-weekly cycling; the paper's full text, not its abstract, is where the tartrate salt is converted to its L-carnitine equivalent

    4.5 g of L-carnitine-L-tartrate daily, equivalent to 3 g of L-carnitine, in a 220 mL beverage containing 44.4 g carbohydrate and 13.8 g protein [8]

  • Randomized, double-blind, placebo-controlled crossover trial in 20 trained male recreational athletes, given as a single acute dose before a high-intensity workout rather than as a loading protocol

    3 g of L-carnitine tartrate 90 minutes before exercise [9]

  • Randomized, double-blind, placebo-controlled trial over 24 weeks in 21 men with idiopathic asthenospermia; the arm combined two different molecules, which is why the trial cannot be read as a test of L-carnitine alone

    2,000 mg of L-carnitine plus 1,000 mg of L-acetyl-carnitine per day [18]

A figure that circulates alongside these and is not one of them: 2,000 mg per day of L-carnitine equivalents, published in 2006 as an observed safe level for chronic supplementation. That is a risk-assessment ceiling derived by two authors affiliated with a supplement-industry trade association, not a dose any trial administered to anyone, and it is reported on this page where it belongs — in the side-effect record — rather than as a protocol range. The approved prescription labeling additionally carries its own titration schedules, which are descriptions of a label rather than guidance and which exist for carnitine deficiency and for dialysis-associated deficiency only — they say nothing about weight, performance, or fertility. None of the ranges above is a recommendation, and none is transferable between forms: doses quoted from an L-carnitine-L-tartrate trial describe the salt, which is not the same mass of L-carnitine, and an intravenous dose given to a patient on dialysis has no bearing on an oral dose in someone whose carnitine pool is already full.

Ranges are what published research reports, not a recommendation. PHL does not prescribe, and nothing here is personalized to you.

Side effects & safety signals

  • Gut-bacterial conversion to trimethylamine and a rise in circulating trimethylamine-N-oxide

    Consistent and diet-dependent. A 24-week randomized, double-blind, placebo-controlled trial in healthy women aged 65 to 70 reported roughly a tenfold rise in plasma trimethylamine-N-oxide on supplementation. · The rise itself is a laboratory finding rather than a symptom, and it is the best-replicated pharmacodynamic consequence of taking L-carnitine. What it means is contested and is set out in the sections above.

    [12] [16]

  • The cardiovascular question attached to that metabolite

    Not established as an effect of supplementation. PubMed searches on 1 September 2026 for L-carnitine with major adverse cardiovascular events or clinical endpoint, restricted to randomized controlled trials, returned no records, and a parallel levocarnitine search with mortality, myocardial infarction or stroke returned two trials, in sepsis and in hemodialysis patients. On that searched record no published trial has tested whether taking L-carnitine changes cardiovascular events in a replete adult. · A meta-analysis of 17 observational studies in 26,167 subjects associated higher circulating trimethylamine-N-oxide with all-cause mortality at a hazard ratio of 1.91, with heterogeneity of 94% on that outcome. A bidirectional Mendelian randomization analysis found no association between genetically predicted trimethylamine-N-oxide or carnitine and coronary artery disease, myocardial infarction, stroke, atrial fibrillation, type 2 diabetes, or chronic kidney disease after correction, while finding that type 2 diabetes and kidney disease raise the metabolite. A 2025 umbrella review of 27 systematic reviews concluded that causal roles remain controversial, that prospective studies in healthy populations were entirely absent, and that confidence in the published syntheses is critically low.

    [13] [14] [15] [16]

  • The upper boundary of characterized long-term safety

    Not an adverse effect but a limit on what has been evaluated. A risk assessment concluded that evidence of safety is strong at intakes up to 2,000 mg per day for chronic supplementation. · The same assessment states that higher intakes have been tested without adverse effects and may be safe, but that the data above 2,000 mg per day are not sufficient for a confident conclusion about long-term safety. Both authors were affiliated with a supplement-industry trade association at the time, which is a reason to read the figure as an industry-derived boundary rather than a regulatory one.

    [19]

  • Sleep disturbance after an acute pre-exercise dose

    Reported by 10% of the 20 trained men in a randomized crossover trial after a single 3 g dose of L-carnitine tartrate. · Minor, self-reported, and from one small trial in which nothing else differed from placebo, including blood pressure. It is a signal worth knowing exists rather than an established effect.

    [9]

  • Tolerability in the largest randomized synthesis

    Across 49 randomized trials in 1,734 adults on maintenance hemodialysis, no adverse effects were reported. · That is reassuring, and it is reassuring about a specific setting: a supervised repletion population whose carnitine is being lost through a dialyser, studied under trial conditions. It is not evidence about long-term use in an adult whose carnitine pool is already full, and a meta-analysis reporting that no adverse effects were reported is a weaker statement than a meta-analysis that measured harm and found none.

    [4]

Published lists are never exhaustive, so report anything unexpected to a licensed provider.

Storage, handling & reconstitution

Lyophilized storage
L-carnitine is not a lyophilized preparation. The three approved prescription dosage forms are an oral tablet, an oral solution, and a ready-to-use injectable solution, and each is kept as the dispensed container's labeling states. Over-the-counter preparations are foods by statute rather than approved drugs, so no manufacturer stability programme or expiry date of the kind that stands behind a prescription product necessarily exists for them, and the container's own labeling is the only stability statement attached to any given one.
Reconstitution
Not applicable. L-carnitine is supplied as an oral tablet, an oral solution, or an injectable solution that is already in solution; none of the approved dosage forms is a lyophilized powder, so there is no diluent, no vial to mix, and no preparation step involved.
Handling notes
The handling question that actually matters for this compound is not temperature but identity: which carnitine, and in which form. L-carnitine, acetyl-L-carnitine, and propionyl-L-carnitine are three different molecules with three different published literatures, and L-carnitine-L-tartrate — the salt most exercise trials administered — is not the same mass of L-carnitine as an equal weight of the free compound. The approved injectable is prescription-only and the approved oral labeling states in its own words that the oral forms are for oral use only and not for parenteral use.

The how-to guides cover the conventions behind these fields, and what the published stability data does and does not establish: Storage, Reconstitution, Handling.

Questions for your provider

Bring this page to a licensed provider. A provider can order labs, review your medications and history, and tell you whether anything here is relevant to your situation. This page can't. Peptide Health Lab does not prescribe and does not sell peptides.

Want a structured starting point for that conversation? The Stack Builder turns your goals into a research-cited outline you can review together.

Citations

20 sources · every identifier checked against PubMed

Before you act on any of this

This page is educational only and is not medical advice. It does not diagnose, treat, or prescribe. Review it with a licensed provider before making any health decision. Peptide Health Lab does not sell peptides.

How we grade and verify evidence · Terms of Use · Privacy Policy

Open the Stack Builder