L-Carnitine for Metabolic Health: Evidence, TMAO, and Safety

L-carnitine helps move fatty acids into the mitochondria, the part of the cell that burns them for energy, which is why it is marketed for energy and weight. The evidence for a metabolic benefit is modest, and there is an open question about what gut bacteria do with it. This guide explains what the pooled trials found, what the TMAO research does and does not show, and how to read a label, as part of the supplements for metabolic syndrome series.

Quick answer: A 2023 pooled analysis of 41 randomized trials, summarized by the NIH Office of Dietary Supplements, found that L-carnitine lowered fasting glucose, insulin resistance, and A1C, but not serum insulin, with large differences between trials. A separate analysis in people with type 2 diabetes reported lower fasting glucose and lower total and LDL cholesterol, but no effect on triglycerides or A1C. A different line of research links carnitine to TMAO, a gut-bacteria by-product associated with cardiovascular risk, though whether supplements cause harm is not established. Common side effects at around 3 grams a day are stomach upset and a fishy body odor.

What carnitine is and where you get it

Your body makes carnitine, and you also get it from food, mostly from red meat and dairy. Healthy people do not usually need extra. Supplements come as L-carnitine, acetyl-L-carnitine (studied mainly for nerve and brain outcomes), and propionyl-L-carnitine (studied for circulation). This guide is about L-carnitine as it is marketed for blood sugar, weight, and metabolic health. Do not assume results from one form carry over to another.

What the research shows

Trials typically ran for weeks to months and used a wide range of doses, so pooled results combine very different protocols. The NIH Office of Dietary Supplements carnitine fact sheet reports the following:

Two points of caution follow from the table. First, the two analyses disagree on A1C, which is a reminder of how much depends on which trials are included. Second, the ODS notes that larger trials are needed to find out whether carnitine lowers the risk of developing diabetes. Pooled averages tell you what happened across a group, not what will happen to you, and lab differences of this size can be small compared with the effect of food and activity changes.

The TMAO question

Gut bacteria can convert carnitine into trimethylamine, which the liver turns into TMAO. In a widely cited 2013 study, researchers examined 2,595 adults undergoing elective cardiac evaluation and also compared people who ate meat with vegans and vegetarians. The ODS summarizes the main findings this way:

• Omnivores produced more TMAO after carnitine than vegans or vegetarians did.
• Higher carnitine levels were associated with cardiovascular disease and events, but only in people who also had high TMAO levels.
• The ODS describes the implications as not well understood.

This is an association, and the authors themselves called for more research on the safety of long-term carnitine supplementation. It does not show that taking a supplement causes heart disease, and it does not show it is safe long term either. For someone who eats a lot of red meat, supplemental carnitine adds to an exposure that is already high. For someone with existing cardiovascular disease, it is a reasonable thing to raise with a clinician before starting. The gut-bacteria angle also connects to the broader question of the microbiome and metabolic health, covered in probiotics and Akkermansia.

A label-math example

Many capsules contain 500 mg of L-carnitine. Reaching 3 grams a day, the level at which the ODS says nausea, vomiting, abdominal cramps, diarrhea, and a fishy body odor can occur, would take 3,000 ÷ 500 = 6 capsules. A trial dose of 2 grams a day for a 12-week study (84 days) adds up to 2 × 84 = 168 grams in total. Neither number tells you what is right for you, but the arithmetic shows how quickly a serving of a few capsules reaches the amounts that cause side effects in trials. If you decide to use it with your clinician, a defined trial period and a marker to follow, such as fasting glucose or A1C, make the result interpretable. A small change in a marker such as the HOMA-IR calculation (fasting glucose in mg/dL times fasting insulin in µU/mL, divided by 405) can be hard to tell from ordinary variation between tests: glucose 100 and insulin 10 gives 2.47, and insulin 9 gives 2.22. Ask your clinician how much change would count as real.

Safety and interactions

• Stomach and body odor. Nausea, vomiting, cramps, diarrhea, and a fishy body odor are reported at about 3 grams a day.
• Seizures. The ODS notes seizures have been reported in people with seizure disorders. If you have one, ask first.
• Kidney disease. Carnitine is cleared differently in kidney disease, and the ODS notes muscle weakness reported in people with uremia. Ask a clinician before using it.
• Medicines that affect carnitine levels. Some anticonvulsants (valproic acid, phenobarbital, phenytoin, carbamazepine) lower blood carnitine, and pivalate-conjugated antibiotics can deplete it with long-term use.
• Upper limit. No tolerable upper intake level has been set, which is not the same as saying any amount is safe.
• Other drugs. Ask a pharmacist to check your full list, because supplement interaction data are incomplete.

Hypoglycemia warning: Combining any blood-sugar-active supplement with insulin, sulfonylureas, or other glucose-lowering medicines can push glucose too low. Symptoms include shakiness, sweating, confusion, and a fast heartbeat. Do not add a supplement to a medication plan without your prescriber’s knowledge.

How it compares with other options

L-carnitine is a lower-tier option compared with choices that have more consistent human data. See the tiers in supplements for insulin resistance, where options such as myo-inositol and magnesium sit above it, and supplements for high triglycerides, where the ODS-reported analysis found no triglyceride effect. If your priority is lowering LDL, supplements for cholesterol covers fiber and sterols, which have steadier results. Carnitine is also sometimes combined with other supplements in weight-loss products, which makes it harder to know what you are taking; see how to read a supplement label.

Choosing a form and a dose range: what the studies used

The ODS summarizes dose ranges used in research for different goals. They are not recommendations; they show how wide the range is and why a product claim of one dose for everything does not make sense.

The point of the table is that the label on a general-purpose carnitine bottle cannot be matched to a single trial. If a product suggests one dose for blood sugar, weight, energy, and heart health, it is borrowing from several unrelated bodies of evidence.

Who might reasonably ask about it, and who should not

People with type 2 diabetes or metabolic syndrome who have already made food and activity changes and are curious about add-ons may reasonably ask a clinician whether carnitine is worth a trial, particularly if they eat little meat. People with seizure disorders, kidney disease, a history of heart disease, or a high red-meat intake have more reasons for caution. People who take anticonvulsants or long-term pivalate-containing antibiotics are a separate group, for whom carnitine status is a medical topic, not a supplement-aisle one.

Common mistakes

• Starting high. Beginning at a dose that already causes stomach upset makes it hard to tell side effects from benefit.
• Mixing up forms. L-carnitine, acetyl-L-carnitine, and propionyl-L-carnitine have different research.
• Using it as a weight-loss shortcut. Average weight effects reported are small.
• Ignoring the diet context. A high red-meat diet already supplies carnitine and may make more TMAO.
• Skipping the plan. Without a marker and an end date, there is no way to judge whether it is worth continuing.

Questions to bring to your appointment

• Is there a reason to expect L-carnitine to help me specifically?
• Do my kidney function or any seizure history make it a poor idea?
• Do my medicines, such as anticonvulsants or antibiotics, change carnitine levels?
• Given the TMAO question, should I avoid long-term use?
• Which marker should we track, and when do we stop if it is not helping?

Frequently asked questions

Does L-carnitine lower blood sugar?

A 2023 pooled analysis of 41 trials reported lower fasting glucose, insulin resistance, and A1C on average, but the trials varied widely in dose and length, and another analysis in people with type 2 diabetes found no A1C effect. Average results are not a prediction for one person.

Does L-carnitine help with weight loss?

Weight-loss trials mostly used 1.8 to 4 grams a day, and any average effect reported is small. It is not a substitute for changes in food, activity, and sleep.

What is TMAO, and should I worry?

TMAO is made when gut bacteria and the liver process carnitine from meat or supplements. A 2013 study linked high carnitine plus high TMAO with more cardiovascular events, but it showed an association, not that supplements cause harm. The ODS calls the implications not well understood. If you have heart disease, ask a clinician before using it.

What are the side effects?

At roughly 3 grams a day, nausea, vomiting, abdominal cramps, diarrhea, and a fishy body odor can occur. Seizures have been reported in people with seizure disorders, and people with kidney disease should ask first.

Is acetyl-L-carnitine the same thing?

No. Acetyl-L-carnitine is studied mostly for nerve and brain outcomes, and results for one form do not carry over to the other.

Can I take it with metformin or insulin?

Carnitine can add to glucose lowering, so combining it with diabetes medicines is a prescriber decision. See the hypoglycemia warning above and supplements and metformin.

About the author: Charles Kirkland is the founder of Metabolic Clarity Labs and author of The Journey Back. He is not a physician. This is general education, not a medically reviewed treatment plan.

Sources checked October 31, 2026. Discuss personal decisions with a qualified clinician or pharmacist. Do not start, stop, or change prescribed treatment based on this article.