Depression is a complex mental disease as a result of sociopsychological, and physiological factors combined. Its main manifestations include disorders of emotion, cognition, behavior, sleep, and appetite (1, 2). The transmission of monoaminergic neurotransmitters and the impairment of neuroplasticity are considered as the core in the pathophysiological mechanism of depression (3).

At present, less than half of the patients with depression are receiving effective treatments, with low cure rate and high recurrence rate (4). Meanwhile, it is still difficult to diagnose depression mainly relying on clinical phenomenology and the judgment of psychiatrists (5). Some studies have found that carnitine supplementation, especially acetyl-L-carnitine (LAC), can significantly improve the symptoms of patients with depression, and is expected to become a new treatment for depression (6). Other clinical studies indicated that the level of LAC in patients with major depressive disorder (MDD) was significantly lower than healthy controls (7). It suggests that carnitine might be involved in the pathological process of depression, which also provides the possibility of carnitine serving as a biomarker for the diagnosis and treatment of depression (7). However, it still remains unknown what types of acylcarnitines (ACs) and their optimal concentrations are helpful for the diagnosis and treatment of depression and other mental illnesses. Systematic studies of this topic are therefore needed for further investigation.

This study mainly reviewed the literature on the determination of ACs in patients with depression or other mental illnesses. It summarized potential ACs as biomarkers of depression and aims to improve our understanding of the carnitine-related pathophysiological mechanism of depression.

Studies were searched in the electronic databases: PubMed, Embase, Cochrane Library, and Web of Science. Moreover, we also searched preprint platform to avoid publication bias. The search was performed by two independent reviewers. Medical Subject-Heading (MeSH) and text words were used as search strategy. And the search strategy was modified according to the database type: (“carnitine” OR “acylcarnitine”) AND (“depression” OR “bipolar disorder” OR “schizophrenia” OR “mood disorders”); (“LAC” OR “Acetyl-L-Carnitine”) AND (“antidepressant” OR “antidepression”) and their matching synonyms were used for an adequate search strategy. Additionally, two independent observers extracted data and resolved differences by discussion to maintain the validity and reliability.

Carnitine (3-hydroxy-4-N-trimethylaminobutyric acid) is a bioactive, water-soluble and low molecular weight amino acid that is usually considered as a dietary supplement (The structural formula of L-carnitine is shown in Figure 1). L-carnitine and D-carnitine isomers exist in mammals. However, only the L-carnitine isomers participate in the lipid metabolism (8). In humans, 75% of L-carnitine derives from diet, while the 25% endogenous part is synthesized from the essential amino acids, lysine, and methionine in the kidneys and liver (9). Furthermore, carnitine homeostasis is maintained via three pathways, namely dietary absorption, limited synthesis rate, and effective renal reabsorption (10).

The hydroxyl group on carnitine can combine with organic acids or fatty acids to form different short-chain, medium-chain, and long-chain esters, namely ACs. ACs and free carnitine (FC) are the main forms of carnitine in human blood (11). ACs are the products of fatty acid β-oxidation, whose composition and concentration changes are also related to the occurrence and development of many diseases, such as congenital metabolic disorders (12), diabetes (13), atherosclerosis (14), tumors (15), and other diseases (The metabolic process of carnitine is summarized in Figure 2).

Carnitine, functioning in the mitochondrial fatty acid oxidation, also plays an important role in regulating the ratio of free coenzyme A (CoA) to acyl-CoA and in the removal of excessive acyl groups (11). In addition, exogenous carnitine derivatives have critical functions in the oxidation of glucose and can improve insulin resistance (16). Carnitine dysfunction may affect the lipid metabolism of mental disorders, suggesting that it may affect the occurrence and development of metabolic syndrome (17). Moreover, the occurrence of carnitine deficiency aggravates the symptoms in some diseases, such as encephalopathy (18), myopathy (19), and cardiomyopathy (20). The symptoms of these diseases are improved after carnitine supplementation (21). For cancer patients, the probability of carnitine deficiency has shown to be very high (22). A research showed that L-carnitine supplementation was safe and might have beneficial effects on fatigue, depressive symptoms and sleep quality in cancer patients (23).

Of the ACs in living organisms, LAC, as the most widely studied endogenous short-chain ACs, has the highest content of ACs and participates in cell anabolism and catabolism pathway (24). It naturally exists in the central nervous system and can travel through the blood-brain barrier (25). ACs have the functions of synthesizing lipids, changing and stabilizing membrane components, regulating gene expressions, improving mitochondrial function, increasing antioxidant activity, and enhancing cholinergic nerve transmission. It promotes acetyl-CoA to enter the mitochondria during fatty acid oxidation, produces acetylcholine, and stimulates the synthesis of protein and membrane phospholipids (26). Furthermore, Acetyl groups provided by LAC and fatty acid β-oxidation can be used to maintain the level of acetyl-CoA (27). LAC also binds with organic acids and free fatty acid fragments, allowing them to be excreted from the cells to prevent them from destroying the cell and tissue structures. In addition to its metabolic activities, LAC also has effects on cytoprotection, antioxidation, and anti-apoptosis in the nervous system (28).

Increasing data seem to prove the hypothesis that carnitine levels are related to the progression of depression (Studies are summarized in Tables 1, 2). Zhang et al. (29) indicated that the average relative intensity of propionylcarnitine (C3:0), and butyrylcarnitine (C4:0) in plasma of overtired and depressed rats was lower than that of healthy control group. Another study found that most medium-chain and long-chain ACs increased in plasma of depressive rats (details are shown in Table 1). After treatments, most of ACs recovered to the normal level, except for the hydroxyoctadecadienoylcarnitine (C18:2-OH), octadecadienylcarnitine (C18:2), eicosadienoylcarnitine (C20:2). The increase of plasma long-chain ACs indicated that the fatty acid oxidation of depressed rats was incomplete, which may lead to a decrease in activities (30). Interestingly, researchers found that the plasma concentrations of carnitine and LAC were significantly higher in the CUMS-induced rat models than those in the control group, but there was no significant difference in the concentration of carnitine and LAC between the fluoxetine treatment group and the model group (31).

Furthermore, the relationship of carnitine and depression has also been explored in different clinical studies. Nasca et al. (7) found that LAC levels in patients with MDD decreased compared with healthy controls, but there was no significant difference in free carnitine. However, another study found that free carnitine was decreased in depression group (32). Moreover, LAC declined more sharply in patients with a history of refractory depression. A history of childhood trauma and emotional neglect (only in women) can be used as the factors to predict the concentration of LAC in refractory depression patients (7). One study explored the differences in plasma metabonomics between MDD patients and healthy people, in which 52 identified metabolites showed differences between the two groups, including the levels of decenoylcarnitine (C10:1), decadienylcarnitine (C10:2), tetradecadienylcarnitine (C14:2), tetradecatrienecarnitine (C14:3), hexanoylcarnitine (C6:0), octanoylcarnitine (C8:0), decanoylcarnitine (C10:0), dodecanoylcarnitine (C12:0), and C3:0. The concentration of above carnitines was negatively correlated with the score of the Hamilton scale (33). Mitro et al. (34) reported that C5:0 and C5:1 were associated with lower odds of antepartum depression whereas C4:0 and C5:0 were with continuous Patient Health Questionnare-9 (PHQ-9) score. Ahmed et al. (35) found that acylcarnitine metabolism can be used to distinguish MDD phenotypes. Free carnitine and short-chain acylcarnitines (C3:0, C4:0, C5-M-DC, C5:1) increased in the core depression (CD+) group. Free carnitine and acylcarnitines (C3:0 and C4:0) increased in the CD+ group and anxious depression (ANX+) group. Acylcarnitines (C8:0, C10:0, and C12:0) decreased in the CD+ group and neurovegetative symptoms of melancholia (NVSM+) group. Acylcarnitines (C5-OH, C16:1, C18:1, C18:2) decreased in all three MDD phenotypes. Furthermore, the acylcarnitine levels of MDD patients changed significantly after treatment. For MDD patients after 2 weeks treatment, LC and LAC levels in the effective treatment group increased significantly, but no significant changes were seen in the ineffective treatment group (36). After 8 weeks of citalopram or escitalopram treatment in MDD patients, the levels of three short-chain acylcarnitines (C3:0, C4:0, and C5:0) increased, and the medium-chain and long-chain acylcarnitines (C8:0, C10:0, C12:0, C14:2, C16:0, C16:1, C18:0, C18:1, and C18:2) decreased (37). In a randomized controlled trial (RCT), Moaddel et al. (38) found that LAC decreased less after ketamine treatment in the MDD group than it did in the placebo group. It demonstrated that low molecular weight compounds in the body can be used as biomarkers for diagnosing and prognosis evaluation of depression. However, some studies have found that the level of carnitine is not significantly correlated with depression. Some patients with low serum carnitine level had no significant improvement after oral administration of L-carnitine (39). Levels of carnitine (total carnitine, free carnitine, ACs) in urine did not differ significantly between healthy controls and patients with depression (40). Jerome et al. (41) found no association between baseline carnitine level and response to S-adenosyl methionine (SAMe) or escitalopram treatments.

Many comorbidities of depression are associated with changes of carnitine levels. Depression is common in hemodialysis (HD) patients (47) which may reduce drug compliance and have a negative impact on the quality of life in HD patients (48). Studies showed that the levels of total carnitine, free carnitine, L-carnitine and C3:0 in dialysis patients were lower than those in healthy controls, while those of C4:0, isoamyl C4:0 and C18:1-OH were higher. Another study also found that acylcarnitine C3:0 was decreased in the end-stage renal disease (ESRD) with depression than those without depression and C3:0 had a good ability to diagnose depression in the ESRD (AUC = 0.934) (42). As long-chain ACs are excreted and metabolized in the kidney, it might be the reason for the increase of serum AC levels in male uremic HD patients. Renal diseases generally lead to L-carnitine homeostasis imbalance as long-term hemodialysis leads to an increase in the ratio of L-AC to free L-carnitine. Such changes interfere with the oxidation of fatty acids and the removal of unwanted short-chain acyl groups from tissues (43, 49). Supplementation of carnitine helps in decreasing self-rating depression scale (SDS) score and improving depressive symptoms in the dialysis group. The state of depression was related to the level of carnitine (44).

Carnitine also plays an important role in other diseases associated with depression, like multiple sclerosis (MS), neuromyelitis optica (NMO), and HIV infection. Decreased levels of total carnitine, free carnitine, and ACs were found in MS and NMO patients. As the most debilitating symptoms, depressive state, and chronic fatigue occur in a majority of patients with MS and NMO (39, 50). Moreover, carnitine deficiency, either from reduced intake or impaired endogenous synthesis, is considered to be the cause of fatigue, which can be treated with oral LAC supplementation (51, 52). Interestingly, there was no difference in carnitine levels between MS patients with or without fatigue. One study revealed that mitochondrial AC deficiency and fatty acid metabolism disorders were not related to fatigue in patients with MS (53). Another study also found no significant correlation between serum carnitine level and scores of three self-evaluation scales in NMO and MS. Determination of serum L-carnitine levels and the benefits of oral administration of L-carnitine remain unclear (39). Depression is also a common complication among patients with HIV infection for their poor treatment outcomes and high mortality (54). Cassol et al. (45) found decreased monoamine metabolites (phenylacetate, 4-hydroxyphenylacetate) and ACs (C3:0, iso C4:0, iso C5:0, 2-methyl C4:0) in plasma distinguished depressed subjects from controls in HIV-positive and HIV-negative cohorts, and these alterations were correlated with the severity of depression. However, a study showed that no significant correlation was found between depression and serum carnitine level as well as dietary carnitine intake (46). In view of the conflicting results presented above, future research should be conducted in larger samples.

Various factors can affect the level of carnitine in vivo. Previous studies have found that age was a factor in determining carnitine level (55, 56). In aged people, the entry of carnitine into cells is negatively affected, thus reducing the metabolism of LAC. Moreover, age has found to be positively correlated with higher AC/FC ratio. Elderly patients with psychosis may suffer secondary carnitine deficiency as higher serum AC/FC ratio is an indicator of secondary carnitine deficiency (21). Besides, the renal reabsorption of LAC may also be impaired, leading to an increase in carnitine level and a decrease in LAC level (57). With regard to gender, there was a significant negative correlation between estrogen level and serum free carnitine in females (58). Therefore, when designing the carnitine determination experiment, in order to avoid the interference of baseline confounding factors, we must pay special attention to the consistency of age, gender, and renal function between the groups, so that the final determination results are comparable.

Bipolar disorder (BD) is a chronic mental illness characterized by severe bipolar changes, ranging from extreme excitement or mania to severe depression. No effective biomarkers have been found for BD, making accurate diagnosis difficult in clinical practice (59). Abnormal mitochondrial function has been found to be present in patients with BD (60). Therefore, several readily-made dietary supplements, including LAC, have been tested as potential treatments for BD with mitochondrial regulatory functions (61, 62). Although LAC supplement is a simple and cost-effective adjuvant therapy used for the treatment of most neurological diseases, its therapeutic effect on BD is controversial (63). Dosage and duration of LAC and α-lipoic acid (ALA) have shown no antidepressant effect on patients with BD, which also had insignificant enhancement of mitochondrial function in patients (62, 64).

A cross-sectional study found that outpatients had higher total carnitine levels in serum than acutely hospitalized schizophrenia and BD patients, but there was no significant difference in total and free carnitine levels of patients with schizophrenia or BD. Serum carnitine is therefore not a reliable marker to distinguish schizophrenia from BD (65). Valproic acid (VPA) is commonly used in several neurological and psychiatric indications. It was found that the dosage of VPA was negatively correlated with the levels of total carnitine and free carnitine, suggesting that sodium valproate had a negative effect on the bioavailability and metabolism of carnitine (55, 66). Another study indicated that the concentration of C2:0 in BD with hyperammonemia and treated with VPA were significantly lower than epilepsy group under phenytoin treatment. Carnitine had a downward trend. This result showed that the level of human blood ammonia was related to the concentrations of carnitine and LAC (56) (Studies are summarized in Table 3).

Carnitine has potential diagnostic and therapeutic effects on schizophrenia (67, 68). Tasic et al. (69) found that C5:0 was present only in schizophrenics, but not in BD and healthy people, suggesting that carnitine was a biomarker to identify schizophrenia. For patients with schizophrenia and metabolic syndrome (MetS), 8 plasma ACs [C3:0, C4:0, C5:0, hexenoylcarnitine (C6:1), C10:1, C10:2, hydroxytetradecadienylcarnitine (C14:2-OH), hydroxyhexadecadienylcarnitine (C16:2-OH)] were significantly higher than those in schizophrenics without MetS (70). Moreover, concentrations of ACs [hydroxybutyrylcarnitine (C4:0-OH) and hexadecenoylcarnitine (C16:1)] in schizophrenia patients were significantly higher than those in healthy controls, whereas those of C3:0, C8:0, C10:0, C10:1, C10:2, C12:0, hydroxytetradecenoylcarnitine (C14:1-OH), C14:2 and C14:2-OH were lower. After treatment, the levels of acetylcarnitine (C2:0), C8:0, C10:0, C10:1, C10:2, C12:0, C12:1, C14:1, C14:1-OH, C14:2, C14:2-OH, C16:2, C16:2-OH, and C18:0 decreased compared to the controls (71). A longitudinal study demonstrated that the levels of C18:0, L-C16:0, C18:2, and LAC decreased in the first-episode and recurrent patients after 8 weeks treatment (72). However, LAC combined with clozapine was found to not significantly improve the treatment effect of schizophrenic patients who were tolerant to clozapine (73) (Studies are summarized in Table 4).

About 2/3 of patients with severe depression are unable to achieve complete remission of symptoms with antidepressants currently available (74, 75). In recent years, there have been a few new and unique antidepressants (76). It has been found that the odds of developing any adverse events are highest for tricyclic antidepressants and serotonin noradrenaline reuptake inhibitors. However, since LAC is an endogenous molecule, it is safer compared with other medicine (77, 78). Unlike some currently used antidepressants that take effect within 2–3 weeks, LAC can take effect within 2–3 days (79). LAC can be used in patients with MDD as a cognitive enhancer (80) and carnitine supplementation is beneficial to depression. LAC supplementation, with the same effect as antidepressants, can significantly reduce depressive symptoms compared with placebo/no intervention (81). Therefore, it is expected to become the next generation of new antidepressants with fewer side effects (79). Previous studies have proposed a variety of mechanisms of LAC in the treatment of depression. Most studied potential mechanisms include neuroplasticity effect, oxidative stress and neurotransmitter regulation. The antidepressant mechanisms of LAC are summarized in Figure 3.

Although many antidepressants with different mechanisms are being used, a number of patients have not achieved appropriate therapeutic effects. The monoamine hypothesis of depression has been the main pathophysiology and drug therapy target of depression over the past few decades, but the current monoamine theory has serious limitations (105). Therefore, new antidepressants with different mechanisms are needed to enable clinicians to diversify the treatment options for MDD. Perhaps LAC may be the next potentially effective and tolerable treatment with new mechanisms for patients with depression, especially for elderly and female patients, as well as patients with concomitant diseases who are vulnerable to the adverse effects of antidepressants. However, compared with other antidepressants, its hypothetical benefits must be thoroughly studied. More large-scale clinical trials are needed to determine whether LAC as a single drug or enhancer is effective and clinically beneficial for the treatment of depression.

The relationship of multiple ACs in vivo and psychiatric disorders such as depression has yet to be further investigated. Most of the current studies on depression have found that the levels of free carnitine and several short-chain acylcarnitines decreased in patients with depression. The changes of medium- and long-chain acylcarnitines are not well-studied but their metabolism may have to do with comorbiditic diseases (43, 49). One Study have found that iso C5:0 exists only in patients with schizophrenia, not in those with bipolar disorder and normal subjects (69), but it still exists in depressive patients (45). Therefore, iso C5:0 may be an important biomarker to distinguish between depression, schizophrenia and BD. Interestingly, it is found that C8:0, C10:0, C10:1, C12:0, and C14:2 have opposite changes in depression and schizophrenia compared with healthy controls (30, 37, 71). The differences in the metabolism of these ACs may potentially become important biomarkers for distinguishing depression and schizophrenia, but these possibilities are yet to be explored in further research (Table 5).

Meanwhile, the determination of carnitine concentrations in serum or plasma did not seem to be significantly different. The comprehensive screening and evaluation of carnitine changes in patients with depression is of great significance to studying the development of depression in depth. It is believed that ACs, especially LAC, might become potential biomarkers for the diagnosis of depression and effective drugs for their treatment of depression in the future given their important clinical value. Comprehensive studies on the relationships of depression and carnitine metabolism can lend support to the diagnosis and treatment of other psychiatric diseases such as BD and schizophrenia, so as to further study the similarities and differences of carnitine metabolism in psychiatric diseases.

TL and KD wrote the manuscript. YX collected the relevant literature and organized it. RuY and RoY checked the content of the manuscript. ZG and MT reviewed and amended the manuscript. All authors read and approved the final manuscript.

The work was supported by National Natural Science Foundation of China (No. 81803233), the Natural Science Foundation of Hunan Province (No. 2018JJ3834), the Key Research Project of Ningxia Hui Autonomous Region in 2021 (Major Project) (2021BEG01001), Science Foundation of Xiangya Hospital for Young Scholar (No. 2017Q13), and China Postdoctoral Science Foundation (No. 2021M693561).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.