L-Carnitine Side Effects: What Actually Happens

The short answer

Most people tolerate oral L-carnitine at ordinary supplement doses. The side effects that do turn up are dose-dependent and mostly unglamorous: nausea, stomach cramps, loose stools, and a distinctive fishy body odour that catches people entirely by surprise.

The odour is the one worth understanding, because it is not an allergy or a bad batch. Gut bacteria convert unabsorbed carnitine into trimethylamine, and trimethylamine smells like old fish. Since only a small fraction of an oral dose is absorbed in the first place, there is a lot of carnitine left in the colon for bacteria to work on, and the odour tracks the dose closely.

Beyond the everyday complaints there are four situations that deserve real caution rather than a shrug: an existing seizure disorder, thyroid treatment, warfarin, and kidney disease. And there is one open question, the TMAO story, that is genuinely unresolved rather than settled in either direction.

The common side effects, and the dose they appear at

The gastrointestinal complaints are the ones people actually report: nausea, vomiting, abdominal cramping and diarrhoea. They cluster at higher doses and improve when the dose is split or lowered. A formal risk assessment of the published trial data put the observed safe level for supplemental L-carnitine at 2,000 mg per day, meaning that was the highest intake at which the trial evidence showed no adverse effects, not a ceiling at which something dangerous begins.

⚠ CLAIM (wellness): Supplemental L-carnitine has an observed safe level of 2,000 mg per day based on published clinical trial data, with gastrointestinal complaints and body odour being the dose-dependent effects reported above that intake. Basis: Hathcock & Shao 2006, Regul Toxicol Pharmacol. Lane: wellness-guidance.

Effect	How common	Typical trigger	What helps
Fishy body, breath or urine odour	Common at higher doses	Roughly 3 g/day and up, sooner in some people	Lower the dose. It reverses on stopping.
Nausea, cramping, diarrhoea	Common	Dose-dependent, worse on an empty stomach	Split the dose, take with food
Restlessness or trouble sleeping	Occasional, mostly with acetyl-L-carnitine	Evening dosing	Move it earlier in the day
Increased seizure frequency	Rare, but reported	Pre-existing seizure disorder	Clinician conversation before starting
Sweating more than usual	Anecdotal, not well characterised	Often reported alongside training	No established mechanism

Why L-carnitine makes you smell like fish

This is the complaint that sends people looking for answers, and the mechanism is specific enough to be worth knowing. Carnitine that is not absorbed passes into the colon, where gut bacteria strip it down to trimethylamine. Trimethylamine is volatile and smells strongly of rotting fish. It leaves the body in breath, sweat and urine, which is why the odour is not fixed by a shower.

Normally the liver oxidises trimethylamine into odourless trimethylamine N-oxide before it reaches your skin. People with the inherited condition trimethylaminuria cannot do this efficiently and develop the odour from ordinary dietary choline and carnitine. Everyone else has a finite capacity, and a large carnitine dose can simply exceed it.

The practical implications follow directly from the mechanism. The odour is dose-related, so lowering the dose usually resolves it. It reverses completely when you stop. It is more likely on the large doses used in fat-loss protocols than on the 500 to 1,000 mg found in a general supplement. And because it depends on your gut bacteria, two people on an identical dose can have completely different experiences, which is why forum advice on this is so contradictory.

The TMAO question, which is genuinely unresolved

The same trimethylamine that causes the odour is the starting point for a much bigger argument about carnitine and heart health, and it is one of the clearest live contradictions in the supplement literature.

⚠ Contradiction: two 2013 papers point in opposite directions on carnitine and cardiovascular risk, and neither has been retracted or resolved. Koeth and colleagues (Nature Medicine) showed that gut bacteria convert dietary carnitine to trimethylamine, which the liver oxidises to TMAO, and that TMAO accelerated atherosclerosis in mice. Omnivores produced far more TMAO from a carnitine challenge than long-term vegans, whose gut flora had largely lost the capacity. Critically, plasma carnitine predicted cardiac risk in their human cohort only when TMAO was also elevated. In the same year DiNicolantonio and colleagues (Mayo Clin Proc) pooled 13 controlled trials covering 3,629 patients and found L-carnitine associated with 27% lower all-cause mortality after myocardial infarction, 65% fewer ventricular arrhythmias and 40% less angina. That analysis drew published criticism for underweighting risk of bias in the older trials it included. Both findings are kept here with provenance rather than resolved.

What can be said fairly is this. The TMAO pathway is real and well demonstrated. Whether raising TMAO through a supplement translates into human cardiovascular events is not established, and the mouse work does not settle it. The trial evidence in cardiac patients points the other way, and those are humans with hard endpoints, though mostly older studies of variable quality.

If you have established cardiovascular disease, this is a conversation for your cardiologist rather than a decision to make from an article. If you are healthy and taking a gram or two a day, there is no evidence of harm and no proof of safety at the level of decades, which is an honest description of most of this category.

The four situations that warrant real caution

An existing seizure disorder

Increases in seizure frequency have been reported in people with pre-existing seizure disorders who took L-carnitine, including in people who had been stable. The mechanism is not well characterised and the reports are not numerous, but the consequence is serious enough that it should not be a solo decision.

⚠ CLAIM (wellness): Increased seizure frequency has been reported in individuals with pre-existing seizure disorders taking L-carnitine. Anyone with a seizure disorder should speak to their neurologist before supplementing. Basis: NIH Office of Dietary Supplements carnitine fact sheet. Lane: wellness-guidance.

Thyroid treatment

This one is genuinely counterintuitive. L-carnitine acts as a peripheral antagonist of thyroid hormone, inhibiting the entry of T3 and T4 into cell nuclei. That property has been deliberately exploited: a randomised trial used it to reduce symptoms in hyperthyroidism, where blunting thyroid hormone action is the goal.

Run the same mechanism in someone being treated for an underactive thyroid and it works against the treatment. If you take levothyroxine, carnitine is a real interaction rather than a theoretical one, and it belongs in a conversation with whoever manages your thyroid.

Warfarin

Case reports describe increased INR and potentiated anticoagulation when carnitine was added to warfarin. Anyone on warfarin adding a supplement should be checking INR more often than usual, and telling the clinic what changed.

Kidney disease and dialysis

People on haemodialysis lose carnitine through the dialysis membrane and can become genuinely deficient, which is why intravenous levocarnitine is an approved prescription treatment in that setting. That is medical care under supervision, not a case for buying a supplement. Reduced kidney function also changes how carnitine and its metabolites clear, so self-supplementing with impaired renal function is not the same proposition as doing it with healthy kidneys.

D-carnitine, and why identity testing matters more than potency here

For most supplements, third-party testing is about whether you are getting the dose on the label and whether it is contaminated with heavy metals. For carnitine there is an additional issue that is specific to this molecule, and it is the more important one.

Carnitine exists as two mirror-image forms. Only L-carnitine is biologically active. The D-isomer is not merely inert: it competitively interferes with L-carnitine transport and can therefore make carnitine status worse rather than better. Racemic DL-carnitine sold in the past was linked to a myasthenia-like syndrome in dialysis patients, which is why the D-form was removed from legitimate supplementation.

So the certificate of analysis you want for carnitine is one that confirms identity and isomeric purity, not just milligrams per serving. A product can hit its stated potency and still be the wrong molecule. Look for a named, verifiable raw material source, a batch-specific certificate of analysis, and third-party verification through a programme such as NSF or Informed Choice. The same due diligence applies to the various salt forms sold as L-carnitine tartrate, fumarate or acetyl-L-carnitine, which differ in how much elemental carnitine they actually deliver per gram.

What people report that the trials do not measure

Two things come up constantly in user reports and are worth naming clearly as anecdote rather than evidence.

The first is sweating more, often noticed during training. There is no established mechanism for this and it is not a documented finding in the trial literature. It may reflect the training rather than the supplement, since carnitine is usually taken by people who are also exercising.

The second is digestive urgency, which does have a plausible basis: an osmotically active compound arriving in the colon largely unabsorbed is a recognisable recipe for loose stools, and it fits the dose-dependence people describe. Users report it, controlled trials record gastrointestinal effects in general terms, and the two are consistent without being the same evidence.

Neither of these has been shown in controlled trials. They are included because they match what people actually search for, clearly labelled as reports rather than findings.

How to lower your risk of running into any of this

Start at the low end, around 500 mg, rather than the 2 to 3 g used in research protocols. Take it with food, and split the dose if you are taking more than a gram. Give it a couple of weeks before deciding whether it does anything, and stop if the odour appears rather than pushing through it.

It is also worth knowing why the large doses exist in the first place. Oral carnitine is only 14 to 18 percent bioavailable against 54 to 87 percent from food, and raising muscle carnitine at all required 2 g twice daily alongside a substantial carbohydrate load for six months in the one protocol that managed it. The high doses in fat-loss protocols are chasing that absorption problem, and the side effects scale with them. Understanding that trade-off is more useful than a list of warnings.

The bottom line

At ordinary doses, L-carnitine’s side effect profile is mild and mostly gastrointestinal, with a fishy odour that is dose-dependent, mechanistically well understood and completely reversible. The situations that deserve genuine care are a seizure disorder, thyroid treatment, warfarin and impaired kidney function. The TMAO question is unresolved and should be held as unresolved rather than as a reason for either alarm or reassurance. And uniquely for this compound, the testing question worth asking a manufacturer is about isomeric identity, not just potency, because the wrong isomer is actively counterproductive rather than simply useless.

Educational information only, not medical advice, and not evaluated by the FDA. L-carnitine is not a treatment for any disease. Talk to a clinician before supplementing if you have a seizure disorder, thyroid disease, kidney disease, cardiovascular disease, are taking warfarin or levothyroxine, or are pregnant or breastfeeding. Side effects described here are drawn from published trial data and safety reporting, and individual responses vary.