Summary of Rationale Behind and Research Studies into L-carnitine as a Treatment for Narcolepsy

SomnusNooze menu
SomnusNooze menu


In this 2013 study published in PLoS One, the researchers Miyagawa et al hypothesized that L-carnitine supplementation might help improve narcolepsy symptoms by increasing β-fatty-acid oxidation.

Fatty acids are a major energy source for the human body. For them to be broken down and metabolized via oxidation and thereby provide vital cellular energy, they must first be transported into the mitochondria (the energy-producing centers in cells) via the carnitine transport system. Carnitine palmitoyltransferase I (CPTI) is the rate-limiting enzyme in this process. This enzyme helps to combine carnitine with a fatty acid to form fatty acyl carnitine, which can then enter the mitochondria through the mitochondria’s membrane.

If this carnitine transport is slowed for any reason, the ability of a cell (presumably, also including brain cells or neurons) to produce energy can be decreased, potentially leading a person to feel more fatigued. For example, if carnitine or CPTI enzyme levels are low, fatty-acid transport and oxidation to produce energy will also be slowed.

Several lines of prior research have implicated problems with carnitine in individuals with narcolepsy (and potentially other primary hypersomnias and even fatigue in general). For example, some people with narcolepsy may have reduced fatty acid oxidation and low levels of acylcarnitine.

Other studies have shown that mice who have a carnitine deficiency have narcolepsy-like symptoms, including disrupted REM sleep, reduced motor activity, and low orexin (also called hypocretin) levels. Orexin is the neuropeptide whose deficiency has been determined to cause narcolepsy with cataplexy (that is, narcolepsy type 1).

Finally, a SNP (single nucleotide polymorphism) has been found in the gene that encodes CPT1B and associates with both narcolepsy type 1 and decreased expression of the CPT1B enzyme itself. SNPs are the most common type of genetic variations within human DNA. They are DNA sequence variations occurring when a single nucleotide (C, G, A, or T) differs between members of paired chromosomes (one pair derived from each parent). In this case, the more C nucleotides (“risk alleles”) a person has at this narcolepsy-related SNP (from 0-2), the lower the production of the CPT1B enzyme. Less CPT1B enzyme leads in turn to slower fatty acid transport and breakdown to energy.

Given the above lines of evidence, Miyagawa et al hypothesized that carnitine supplementation might help improve symptoms of narcolepsy type 1 by increasing the transport of fatty acids into mitochondria for energy production.


The study design was a randomized, double-blind, placebo-controlled cross-over trial. The researchers studied 30 people with narcolepsy with cataplexy, of which 28 completed the trial. Each person underwent two 8-week treatment periods, one with the placebo and one with L-carnitine (the order of which was randomly determined). Dosage was 2 capsules in the morning and 1 capsule in the evening. Capsules contained either 170 mg L-carnitine or placebo. Participants were monitored every 4 weeks, and they were allowed to continue their other usual medications.

“The primary outcome measure was the patients’ subjective assessment of their sleepiness using total time for dozing off during the daytime in their sleep logs.” Secondary outcome measures (meaning that they were not the main purpose of the study) included:

  1. The number of occurrences of daytime dozing off
  2. The number of episodes of sleep paralysis and cataplexy, according to sleep logs
  3. Japanese version of the Epworth Sleepiness Scale
  4. SF-36 (Medical Outcomes Study 36-Item Short-Form Health Survey): vitality and mental health subscales
  5. Body mass index


Total time for dozing off during the daytime (the primary outcome) was significantly reduced in participants while ingesting L-carnitine, as compared with when ingesting the placebo. No significant differences were found in any of the secondary outcomes measured. No participants experienced side effects when ingesting L-carnitine.

About half the participants had at least 1 narcolepsy-associated risk allele in the gene encoding CPT1B. However, the number of risk alleles (for example, 0, 1, or 2) did not predict response in overall nap times between L-carnitine and placebo. In layman’s terms, the presence of the higher-risk allelic variant did not seem to affect whether or not L-carnitine supplementation improved sleepiness.

Overall, the participants in this study had relatively low levels of acylcarnitine in their blood while taking placebo. After taking L-carnitine, their acylcarnitine levels improved significantly, as did their total and free carnitine levels.


These results support the hypothesis that taking L-carnitine supplements improves sleepiness or dozing in people with narcolepsy type 1. These improvements occurred regardless of whether or not people had a baseline carnitine deficiency.

Although this study was, in part, motivated by the finding of a gene association (SNP), the gene association did not predict efficacy of L-carnitine supplementation. L-carnitine supplementation showed efficacy in this modest-sized group of participants regardless of their CPT1B narcolepsy risk allele status. That being said, the study was not designed or “powered” to assess the effect of the gene on improvements with L-carnitine.

This study was also not powered (meaning that the number of participants was insufficient to show a statistical difference between the L-carnitine and placebo groups) for the secondary endpoints. Some of these secondary endpoints, in fact, did show some tendency towards improvement (see, #1, #3, and #4 above). Cataplexy and sleep paralysis improvements could not be well studied because patients were allowed to continue taking their regular medications during this study and subsequently had very low levels of cataplexy and sleep paralysis. This is sometimes referred to as a “floor effect.” That is, the absence of evidence (e.g., a desirable or suspected positive benefit to an intervention) is not necessarily evidence of absence. This is also referred to as Type II error (or a potential false-negative finding) because not enough observations were made or the choice of what to observe (e.g., cataplexy events) is so suppressed that any change in the number of events is artificially dampened (and thereby elude detection).

Participants in this study were found to have reduced triglyceride levels (though not significantly), indicating that carnitine absorption and use was likely normal in these people with narcolepsy type 1. It is possible that longer-term use of L-carnitine might improve body mass index and lipid profiles in people with narcolepsy. However, studies to assess these outcomes usually require 6 months to 4 years or more.

Further Reading

T Miyagawa, Kawamura H, Obuchi M, et al. Effects of oral L-carnitine administration in narcolepsy patients: a randomized, double-blind, cross-over and placebo-controlled trial. PloS One. 2013;8(1):e53707. The full text of this open-access article is available at

This article was written for the Hypersomnia Foundation by a volunteer medical writer and reviewed by Dr. David Rye, Chairperson, HF Scientific Advisory Board.

Share Post

SomnusNooze homepage