In short: intravenous levocarnitine is a licensed prescription medicine with two approved indications — inborn errors of metabolism causing secondary carnitine deficiency, and carnitine deficiency in end-stage renal disease on dialysis — and its own label states that in the dialysis population the effects on the signs, symptoms and clinical outcomes of that deficiency have not been determined. Everything else is either unproven or belongs to a different route. Skeletal muscle carnitine uptake runs through a sodium-dependent transporter that is saturated at normal plasma concentrations, so an infusion raises blood carnitine without loading the tissue; the only human study to move muscle carnitine did so by combining a five-hour infusion with clamped high insulin, and gained about 12%. The weight-loss data — 37 randomised trials, 2,292 participants, −1.21 kg body weight — are entirely oral and confined to participants with overweight or obesity. The cardiac data split by trial size: a pooled 27% mortality reduction across 13 mostly pre-reperfusion-era trials, against a neutral primary endpoint in CEDIM-2, the largest dedicated trial. No trial supports intravenous L-carnitine for fat loss, athletic performance or fatigue in a healthy adult.
Carnitine arrives in the infusion clinic with an unusually clean origin story. It is not a botanical extract or a repurposed antioxidant; it is a molecule with a defined transport job, a licensed injectable form, a biochemically defined deficiency state and a diagnostic threshold printed on a regulatory label. That is more than most compounds in this category can claim. What it does not have is any evidence that giving it intravenously to someone who is not deficient changes anything — and the reason for that is mechanical rather than statistical.
What "L-carnitine IV therapy" covers
Three practices share the name. The first is levocarnitine as a prescription drug: gram-scale dosing for inborn errors of metabolism, given as a slow bolus or infusion with plasma monitoring, and a separate dialysis regimen given into the venous return line. In the United Kingdom it is stocked as Carnitor 1 g/5 mL injection ampoules; in the United States the same product carries an FDA-approved label.
The second is its use in acute toxicology — most established in valproate-induced hyperammonaemia and hepatotoxicity, where carnitine depletion is part of the mechanism of injury. This is a rescue indication with a real rationale and an unsettled evidence base, not a wellness one.
The third is the metabolic or "fat-burner" drip: L-carnitine as a component of a multi-ingredient infusion, marketed for fat oxidation, energy, recovery or metabolic support in healthy adults. That third practice is where almost all commercial interest sits, and it is the one with no dedicated outcome trials at all.
Mechanism: the carnitine shuttle
Carnitine's function is transport, not fuel. Long-chain fatty acids cannot cross the inner mitochondrial membrane as acyl-CoA. Carnitine palmitoyltransferase 1 on the outer membrane transfers the acyl group onto carnitine; carnitine-acylcarnitine translocase carries the resulting acylcarnitine across the inner membrane; carnitine palmitoyltransferase 2 hands the acyl group back to CoA on the matrix side, where beta-oxidation proceeds. A second, quieter role matters just as much in disease: the same shuttle exports accumulating acyl groups out of the mitochondrion, which is why carnitine is consumed in conditions that generate abnormal acyl compounds.
Carnitine is conditionally essential rather than essential. Roughly three-quarters of body carnitine comes from diet — chiefly red meat and dairy — and the remaining quarter is synthesised endogenously from protein-bound lysine and methionine. That synthesis is not a standalone pathway: it depends on iron, on pyridoxal 5'-phosphate, on NAD and on ascorbate as cofactors, and it releases glycine as a by-product of one of its steps. This is the honest reason carnitine appears alongside lysine, vitamin C and B-complex vitamins in formulation thinking — they are upstream of it — and it is not, on its own, evidence that any of them raise carnitine status in a person eating an ordinary diet.
The uptake ceiling that decides everything
About 99% of body carnitine is intracellular, with the highest concentrations in cardiac and skeletal muscle, held against a steep gradient by the sodium-dependent transporter OCTN2. That gradient is the entire problem for infusion therapy. The transporter is saturated at normal plasma concentrations, so pushing plasma concentrations higher does not increase net transport into muscle. Repeated attempts to raise muscle carnitine content in healthy humans by oral or intravenous dosing failed for precisely this reason.
One experiment defines the boundary. Eight healthy men received a five-hour intravenous L-carnitine infusion while serum insulin was held at a high physiological concentration, around 149 mIU/L, by clamp. Under those conditions — and only under those conditions — muscle total carnitine rose from 22.0 to 24.7 mmol per kg dry mass, accompanied by a 2.3-fold increase in OCTN2 transporter messenger RNA. Hypercarnitinaemia alone did nothing; hypercarnitinaemia plus hyperinsulinaemia moved the tissue by roughly a tenth.
Read that result the way a formulator should. It is a proof of mechanism, not a protocol: the insulin concentrations required are supraphysiological in the fasted state and are produced in a research setting, not in an infusion chair. What it establishes is the direction of the constraint. In a healthy adult, delivery is not the rate-limiting step for carnitine — transport is — and an intravenous route solves the wrong problem. That is the mirror image of the argument for high-dose vitamin C, where the intravenous route genuinely reaches concentrations oral dosing cannot.
Where the route does matter is absorption in the deficient patient. Oral bioavailability of supplemental L-carnitine is in the order of 14 to 18% of the dose, while the intravenous route bypasses that limitation entirely and the kidney reclaims the great majority of what it filters. When the goal is to restore a depleted pool rather than to load a healthy one, that difference is real.
Who is actually carnitine deficient
The label sets a number, which is more than most infusion ingredients offer. Carnitine deficiency is defined biochemically as a plasma free carnitine concentration below 20 µmol/L, or an acylcarnitine-to-levocarnitine ratio above 0.4, or abnormally elevated urinary acylcarnitine.
Primary carnitine deficiency is a genetic defect of the OCTN2 transporter itself, caused by mutations in SLC22A5, with a reported frequency of roughly one to five per 10,000 and tissue concentrations that can fall below 10% of normal. Secondary deficiency is far more common and always has a cause behind it: inborn errors of metabolism that trap carnitine as acylcarnitines, haemodialysis, valproate therapy, and some malabsorptive states. Strict vegetarians and vegans have lower intakes and rely more on endogenous synthesis, but this does not by itself produce the deficiency state the label describes.
The practical consequence is that carnitine deficiency, unlike the magnesium or B12 status questions covered elsewhere in this Journal, is measurable. A plasma free carnitine and an acylcarnitine profile answer the question directly. An infusion given without them is treating an assumption.
Dialysis: the licensed indication, read honestly
Haemodialysis removes carnitine efficiently, and long-term dialysis patients develop a genuine, measurable depletion. On that basis, injectable levocarnitine was approved for the prevention and treatment of carnitine deficiency in end-stage renal disease patients undergoing dialysis, at 10 to 20 mg/kg dry body weight given as a slow two-to-three-minute bolus into the venous return line after each session.
The label then does something unusual, and it deserves to be quoted rather than paraphrased: in this population, "the effects of supplemental carnitine on the signs and symptoms of carnitine deficiency and on clinical outcomes have not been determined." The approval is for correcting a laboratory abnormality, not for improving how patients feel or fare.
The trial literature has not closed that gap. A systematic review and meta-analysis of 49 randomised trials in 1,734 haemodialysis patients found that L-carnitine reduced LDL cholesterol and C-reactive protein, with no significant difference in triglycerides, total cholesterol, HDL cholesterol, haemoglobin, haematocrit, albumin or the required erythropoietin dose. Later meta-analyses have reported reductions in intradialytic hypotension and signals on cardiac function, but the trials are small, heterogeneous in route and dose, and short. This is the most evidence-supported use of intravenous carnitine that exists — and it is still an indication where the regulator declined to claim a clinical benefit.
The cardiac evidence and its size problem
Carnitine's cardiac literature follows a pattern this Journal has now documented in several molecules: encouraging pooled results from many small trials, and a neutral result from the largest single one.
The pooled case is a 2013 meta-analysis in Mayo Clinic Proceedings of 13 controlled trials in 3,629 patients after myocardial infarction, reporting a 27% reduction in all-cause mortality, a 65% reduction in ventricular arrhythmias and a 40% reduction in angina. Correspondence published in the same journal disputed it directly, arguing that the risk of bias had been underestimated and that the body of evidence was of very low quality — much of it generated before routine reperfusion therapy, in an era when the metabolic hypothesis had more room to operate.
The counterweight is CEDIM-2, which tested the hypothesis prospectively in acute anterior ST-elevation myocardial infarction: 9 g per day intravenously for five days, then 4 g per day orally for six months. The primary composite of death and heart failure at six months was 9.2% with carnitine against 10.5% with placebo, a non-significant difference (p=0.27). Five-day mortality, a secondary endpoint, was lower with carnitine (HR 0.61, 95% CI 0.37–0.98). The trial stopped at 2,330 patients of a planned 4,000 because of slow enrolment, which makes it underpowered rather than conclusive — the honest reading is an unresolved question, not a refutation and not a licence.
Fat loss, performance and the oral literature
The commercial claim for a carnitine drip is fat oxidation. The evidence cited in support of it is almost entirely oral, and it is more modest than the marketing implies.
A meta-analysis of 37 randomised controlled trials in 2,292 participants found L-carnitine supplementation reduced body weight by 1.21 kg, body mass index by 0.24 kg/m² and fat mass by 2.08 kg, with no significant change in waist circumference or body fat percentage. A dose-response analysis identified about 2,000 mg per day as the point of maximum effect, and subgroup analysis found the effect only in participants with overweight or obesity. Restricting the pooling to high-quality trials narrowed the confirmed effect to body weight alone.
In exercise, meta-analyses of oral dosing at 1 to 3 g per day report reduced markers of exercise-induced muscle damage — creatine kinase and myoglobin — and less delayed-onset soreness at 24 and 48 hours, while performance results across individual trials remain mixed. In idiopathic male infertility, a systematic review of eight randomised trials found improvements in sperm motility and morphology, no effect on sperm concentration, and no demonstrable effect on pregnancy outcomes in the few trials that reported them — with high heterogeneity and low-quality underlying evidence, as the authors themselves state. All of it is oral, over weeks. None of it transfers to a single intravenous session, and the uptake data explain why it would not be expected to.
The TMAO question, and what the route changes
In 2013 a study in Nature Medicine reported that gut microbiota metabolise dietary L-carnitine to trimethylamine, which the liver oxidises to trimethylamine-N-oxide, and that this pathway accelerated atherosclerosis in mice; in humans, plasma carnitine predicted cardiovascular events only among subjects who also had high TMAO. It was an elegant result and it reframed carnitine as a potential cardiovascular liability.
The causal claim has not consolidated. Mendelian randomisation studies have generally not supported a causal role for TMAO in cardiovascular disease, and observational findings conflict across populations — including in populations with high dietary TMAO intake and low atherosclerosis burden. The mechanism remains a legitimate area of study; the risk attribution does not currently support a clinical rule.
Two things can be stated without overreaching. The pathway is microbiota-dependent, which means the intravenous route bypasses the gut conversion step that generates trimethylamine in the first place. And the one place a regulator has written the concern into a label is narrow and specific: high-dose oral levocarnitine in patients with severely compromised renal function or end-stage renal disease may cause accumulation of trimethylamine and trimethylamine-N-oxide.
Clinical evidence graded by indication
Set side by side, carnitine's evidence is strongest in inherited and dialysis-associated deficiency, unsettled in cardiology and toxicology, and absent in the indications that sell.
| Indication | Evidence base | Key finding | Verdict |
|---|---|---|---|
| Inborn errors causing secondary carnitine deficiency | FDA-approved indication; label dosing 50 mg/kg | Approved for acute and chronic treatment; plasma monitoring to 35–60 µmol/L | Established — approved indication |
| Carnitine deficiency in ESRD on dialysis | FDA-approved indication; 49-RCT meta-analysis, n=1,734 | Corrects the biochemical deficiency; LDL and CRP reduced; no effect on haemoglobin, triglycerides, total or HDL cholesterol. Label: clinical outcomes "have not been determined" | Approved for repletion — outcomes unproven |
| Primary carnitine deficiency (SLC22A5/OCTN2) | Genetic disorder, 1–5 per 10,000; lifelong supplementation standard | Tissue carnitine can fall below 10% of normal; repletion is the treatment | Established — disease-specific |
| Valproate-induced hyperammonaemia and hepatotoxicity | Case series, historical-control comparisons, one negative cohort | Hepatic survival 47% vs 10% against historical controls; a later poisoning cohort found no improvement in management. Route and dose unsettled | Plausible rescue use — evidence unsettled |
| Secondary prevention after myocardial infarction | Meta-analysis of 13 trials, n=3,629; CEDIM-2 RCT, n=2,330 | Pooled: mortality −27%, arrhythmias −65%, angina −40%, disputed as very low quality. CEDIM-2 primary endpoint 9.2% vs 10.5% (p=0.27); day-5 mortality HR 0.61 | Contested — largest trial neutral |
| Muscle carnitine loading in healthy adults | Mechanistic human study, n=8, insulin clamp | Infusion alone does not raise muscle carnitine; with clamped hyperinsulinaemia, 22.0 → 24.7 mmol/kg dm | Not achievable by infusion alone |
| Weight and fat mass reduction | Meta-analysis, 37 RCTs, n=2,292 — oral dosing | −1.21 kg weight, −2.08 kg fat mass, −0.24 kg/m² BMI; no change in waist circumference or body fat %; effect limited to overweight/obesity | Modest and oral — not an IV indication |
| Exercise recovery and performance | Meta-analyses of oral 1–3 g/day | Reduced creatine kinase, myoglobin and soreness at 24–48 h; performance results mixed | Limited — surrogate markers, oral route |
| Idiopathic male infertility | Systematic review, 8 RCTs — oral carnitine/acetyl-carnitine | Improved motility and morphology; no change in concentration; no demonstrable effect on pregnancy outcomes; low-quality evidence | Surrogate benefit — outcome unproven |
| Fatigue, metabolic optimisation, wellness infusion | No dedicated outcome RCTs identified | — | Evidence not established |
Dosing, route and safety
Licensed intravenous dosing is specific. For metabolic disorders the recommended dose is 50 mg/kg as a slow two-to-three-minute bolus or by infusion, with a loading dose in severe metabolic crisis and subsequent dosing every three to six hours, titrated against plasma concentrations in a 35–60 µmol/L target range. For dialysis patients it is 10 to 20 mg/kg dry body weight after each session, with downward adjustment guided by trough concentrations from around week three or four. Both regimens assume a documented deficiency and a monitoring plan.
Safety is not the reason to be cautious with carnitine, but it is not nothing either. The label warns of serious hypersensitivity reactions, including anaphylaxis, laryngeal oedema and bronchospasm, reported mostly in dialysis patients. Seizures have been reported post-marketing, in patients with and without pre-existing seizure disorders. Levocarnitine has been reported to raise the INR in patients on warfarin, so monitoring is advised when it is started, stopped or adjusted. Safety and efficacy have not been evaluated in renal insufficiency outside the dialysis indication. Injection-site reactions, headache, nausea, vomiting and body odour are the commonly reported trial effects; the odour is trimethylamine, and it is a direct readout of the microbial pathway discussed above.
How to read a carnitine formulation
Three questions do most of the work. First, deficiency or optimisation — is there a plasma free carnitine and acylcarnitine profile behind this infusion, or an assumption? Carnitine is one of the few molecules in this category where the question is answerable by a blood test, which raises rather than lowers the standard. Second, route — is the evidence being cited oral or intravenous? Almost all of the weight, performance and fertility literature is oral, over weeks, and the uptake ceiling is a reason not to expect it to transfer. Third, claim — is the formulation described as supporting fatty-acid transport and metabolic balance, or as causing fat loss? Those are different statements, and only the first is supportable.
Carnitine's position in a multi-ingredient concept is defensible on physiology: it is a required transport cofactor with a well-characterised shuttle, a licensed injectable form and a measurable status marker, and it sits downstream of nutrients a formulation may already contain. That is the basis on which it appears in a metabolic concept such as Carnizin, paired with zinc, and it is a reasonable one. It is not a basis for importing the dialysis approval, the pooled cardiac meta-analysis or the oral weight-loss trials as evidence for a product. Keeping those apart — what the molecule is known to do, what a route can deliver, and what a given formulation has been shown to do — is the standard EFBA applies. With carnitine the discipline is unusually easy to apply, because the physiology draws the line before the marketing gets a chance to.
Frequently asked questions
What is intravenous L-carnitine actually approved for?
Two indications, both of them deficiency states. Levocarnitine injection is approved for the acute and chronic treatment of patients with an inborn error of metabolism resulting in secondary carnitine deficiency, and for the prevention and treatment of carnitine deficiency in patients with end-stage renal disease undergoing dialysis. The label defines that deficiency biochemically — plasma free carnitine below 20 µmol/L, or an acylcarnitine to levocarnitine ratio above 0.4. Fat loss, athletic performance, fatigue and general metabolic optimisation are not approved indications, and no regulator has assessed the product for them.
Does an L-carnitine infusion increase carnitine in muscle?
Not on its own. Around 99% of body carnitine is intracellular, and skeletal muscle takes it up against a steep concentration gradient through the sodium-dependent OCTN2 transporter, which is already saturated at normal plasma concentrations. Repeated attempts to raise muscle carnitine by oral or intravenous dosing failed for that reason. The exception is instructive: when eight healthy men received a five-hour intravenous L-carnitine infusion while serum insulin was clamped at a high physiological concentration, muscle total carnitine rose from 22.0 to 24.7 mmol per kg dry mass, with a 2.3-fold increase in OCTN2 transporter mRNA. Delivery is not the bottleneck in a healthy adult — transport is.
Does IV L-carnitine cause fat loss?
No trial has tested that. The weight-related evidence base is oral, not intravenous. A meta-analysis of 37 randomised trials in 2,292 participants found L-carnitine supplementation reduced body weight by 1.21 kg, BMI by 0.24 kg/m² and fat mass by 2.08 kg, with no significant change in waist circumference or body fat percentage; the effect was confined to participants with overweight or obesity, and the dose-response analysis pointed to about 2,000 mg per day as the ceiling. Those are small average changes over weeks of daily oral dosing, not the mechanism a single infusion is marketed on. The physiological reason is the same one above: carnitine is a transport cofactor, not a rate-limiting fuel, and adding more to plasma does not increase the amount inside muscle.
What did the cardiac trials of L-carnitine actually find?
They disagree, and the disagreement follows trial size. A 2013 meta-analysis in Mayo Clinic Proceedings pooled 13 controlled trials in 3,629 patients after myocardial infarction and reported a 27% reduction in all-cause mortality, a 65% reduction in ventricular arrhythmias and a 40% reduction in angina. Correspondence in the same journal argued that the risk of bias had been underestimated and the underlying evidence was of very low quality, much of it predating modern reperfusion therapy. The largest dedicated trial, CEDIM-2, randomised 2,330 patients with acute anterior ST-elevation myocardial infarction to 9 g per day intravenously for five days followed by 4 g per day orally for six months: the primary composite of death and heart failure at six months was not significantly different (9.2% versus 10.5%, p=0.27), with a reduction in five-day mortality as a secondary endpoint (HR 0.61, 95% CI 0.37–0.98). It was stopped early at 2,330 of a planned 4,000 patients, so it is underpowered rather than definitive.
Is IV L-carnitine safe, and who should avoid it?
The label carries a hypersensitivity warning: serious reactions including anaphylaxis, laryngeal oedema and bronchospasm have been reported, mostly in dialysis patients. Seizures have been reported after marketing, in patients with and without pre-existing seizure disorders. Levocarnitine has been reported to increase the INR in patients on warfarin, so monitoring is advised when it is started, stopped or adjusted. Safety and efficacy have not been evaluated in renal insufficiency outside the dialysis indication, and high-dose oral dosing in severely impaired renal function can allow accumulation of the microbial metabolites trimethylamine and trimethylamine-N-oxide. Injection-site reactions, headache, nausea, vomiting and body odour are the commonly reported effects in trials.
Is the carnitine-TMAO cardiovascular concern settled?
No. A 2013 study in Nature Medicine showed that gut microbiota convert dietary L-carnitine to trimethylamine, which the liver oxidises to trimethylamine-N-oxide, and that this accelerated atherosclerosis in mice; in humans, plasma carnitine predicted cardiovascular events only in those with concurrently high TMAO. The causal step has not held up cleanly since — Mendelian randomisation studies have generally not supported a causal role for TMAO in cardiovascular disease, and observational data conflict across populations. Two things can be said with confidence: the pathway is microbiota-dependent, so the intravenous route bypasses the gut conversion step that generates it, and the regulatory warning about TMA and TMAO accumulation is written specifically about high-dose oral dosing in severe renal impairment.
Work with clinically-grounded formulations
EFBA partners with clinicians, pharmacists and distributors across the Arab world on science-driven anti-aging and longevity concepts. The IVIXIR series is formulated to professional-grade standards — with the same evidence discipline applied here.
Selected references
- CARNITOR (levocarnitine) injection, solution — US prescribing information. DailyMed, National Library of Medicine. dailymed.nlm.nih.gov
- Levocarnitine injection, solution — US prescribing information (warnings, precautions, dosage). DailyMed, National Library of Medicine. dailymed.nlm.nih.gov
- Flanagan JL, Simmons PA, Vehige J, Willcox MDP, Garrett Q. Role of carnitine in disease. Nutr Metab (Lond). 2010;7:30. pmc.ncbi.nlm.nih.gov
- Stephens FB, Constantin-Teodosiu D, Laithwaite D, Simpson EJ, Greenhaff PL. Insulin stimulates L-carnitine accumulation in human skeletal muscle. FASEB J. 2006;20(2):377–379. faseb.onlinelibrary.wiley.com
- Stephens FB, Constantin-Teodosiu D, Greenhaff PL. New insights concerning the role of carnitine in the regulation of fuel metabolism in skeletal muscle. J Physiol. 2007;581(2):431–444. pubmed.ncbi.nlm.nih.gov
- Chen Y, Abbate M, Tang L, et al. L-Carnitine supplementation for adults with end-stage kidney disease requiring maintenance hemodialysis: a systematic review and meta-analysis. Am J Clin Nutr. 2014;99(2):408–422. pubmed.ncbi.nlm.nih.gov
- DiNicolantonio JJ, Lavie CJ, Fares H, Menezes AR, O'Keefe JH. L-Carnitine in the secondary prevention of cardiovascular disease: systematic review and meta-analysis. Mayo Clin Proc. 2013;88(6):544–551. pubmed.ncbi.nlm.nih.gov
- Regarding L-carnitine and cardiovascular disease (correspondence) and authors' reply. Mayo Clin Proc. 2013. mayoclinicproceedings.org
- Tarantini G, Scrutinio D, Bruzzi P, Boni L, Rizzon P, Iliceto S. Metabolic treatment with L-carnitine in acute anterior ST segment elevation myocardial infarction: a randomized controlled trial (CEDIM-2). Cardiology. 2006;106(4):215–223. pubmed.ncbi.nlm.nih.gov
- Talenezhad N, Mohammadi M, Ramezani-Jolfaie N, Mozaffari-Khosravi H, Salehi-Abargouei A. Effects of l-carnitine supplementation on weight loss and body composition: a systematic review and meta-analysis of 37 randomised controlled clinical trials with dose-response analysis. Clin Nutr ESPEN. 2020;37:9–23. pubmed.ncbi.nlm.nih.gov
- Koeth RA, Wang Z, Levison BS, et al. Intestinal microbiota metabolism of L-carnitine, a nutrient in red meat, promotes atherosclerosis. Nat Med. 2013;19(5):576–585. pubmed.ncbi.nlm.nih.gov
- Gziut T, Thomas SHL, Eddleston M, et al. L-carnitine for valproic acid-induced toxicity. Br J Clin Pharmacol. 2025;91(2):283–296. pmc.ncbi.nlm.nih.gov
- Vodovar D, Beaune S, Langrand J, et al. L-carnitine does not improve valproic acid poisoning management: a cohort study with toxicokinetics and concentration/effect relationships. Clin Toxicol. 2022. ncbi.nlm.nih.gov
- Khaw SC, Wong ZZ, Anderson R, Martins da Silva S. l-carnitine and l-acetylcarnitine supplementation for idiopathic male infertility. Reprod Fertil. 2020;1(1). ncbi.nlm.nih.gov