Feline hepatic lipidosis (FHL) is defined by the excess storage of hepatic lipids and is considered the most common liver disease affecting cats in North America (1). Although the exact pathophysiology of FHL has not been identified, there is evidence that the mobilization of free fatty acids from the adipose tissue leads to the accumulation of lipids within the livers of affected cats (2). Food deprivation, whether due to secondary anorexia or due to an imposed high degree of energy restriction, is considered a key factor in the development of FHL (1, 3–6). Left untreated, the prognosis for affected cats is considered poor and may result in liver failure and death (7). Obese cats are at an increased risk of developing FHL due to their increased levels of adipose tissue and subsequent stores of fatty acids available for mobilization (2, 8). Also, higher concentrations of hepatic triglycerides (TAG) and an increased risk of concurrent insulin resistance are believed to leave obese cats more susceptible to FHL (9, 10). Additionally, energy restriction is often recommended in obese cats in order to lose weight. However, a low degree of dietary energy restriction often does not result in weight loss in an obese cat, and therefore may result in a higher degree of energy restriction may be implemented (11, 12).

Under normal conditions, fatty acids within the hepatocytes can undergo β-oxidation within the mitochondria for the production of acetyl-CoA, or they can be re-esterified to TAG (13, 14). These TAG can be secreted into the circulation or they can continue to accumulate within the hepatocytes. It is unclear what leads to the accumulation of hepatic TAG during FHL. However, the metabolic idiosyncrasies that result from their obligate carnivorous nature put cats at risk for deficiencies in certain dietary essential nutrients. Insufficient intakes of dietary choline and carnitine during energy restriction have been proposed as potential risk factors in the pathogenesis of FHL (15).

Choline is a precursor for the biosynthesis of phosphatidylcholine (PC). Specifically, PC is required for packaging TAG and cholesterol into very low-density lipoproteins (VLDL) for export out of the liver (16). Betaine, a derivative of choline, also serves as a methyl group donor for one of the pathways through which homocysteine is remethylated to methionine. The production of methionine subsequently results in the methylation and synthesis of numerous metabolites, including carnitine (17). Previous research in mammals, including cats, found that choline deficiency leads to the development of fatty liver, which can be reversed and/or prevented when dietary choline is added back to the diet in sufficient quantities (18–20). In cats, increased dietary choline supplementation leads to increased concentrations of serum lipids and lipoproteins in both obese and overweight cats (21, 22), suggesting increased lipid hepatic mobilization through the synthesis and secretion of VLDL from the liver. Additionally, choline and its derivative betaine may reduce fat mass gain, as suggested in livestock species and growing kittens fed ad libitum post-gonadectomy (23–25).

In comparison, carnitine is required to facilitate the entry of fatty acids into the mitochondria for β-oxidation. This is considered the main pathway for the disposal of fatty acids under normal physiological conditions and is important for the production of ATP by the mitochondria (13). L-carnitine increases lipid oxidation in overweight cats, as suggested by increased palmitate oxidation (26), energy expenditure (EE) and lower post-prandial respiratory quotient (RQ) (27). Additionally, L-carnitine supplementation may decrease food intake and weight gain in ad libitum feeding scenarios (28). For these reasons, L-carnitine is commonly used by the pet food industry in the formulation of weight control and weight loss diets for cats. In human medicine, reduced β-oxidation has been suggested as a potential cause of lipid accumulation in the liver of patients with non-alcoholic fatty liver disease (NAFLD) (29). Thus, it has been proposed that an insufficient supply of carnitine may similarly be an important factor for the development of FHL, although the evidence to support this has been inconsistent (30–34).

Nutrients and nutraceuticals that can reduce the risk of hepatic lipid accumulation in obese cats and assist in mitigating weight gain would benefit the pet food industry and the veterinary community. Currently, there is evidence that dietary choline may provide such benefits to overweight and obese cats (21, 22). However, there have been no studies comparing the lipotropic benefits of dietary choline to L-carnitine, a dietary supplement commonly used by the pet food industry. Therefore, the objective of this research was to investigate the lipotropic effects of dietary choline compared to L-carnitine (a positive control) and no supplement (a negative control), on serum lipid and lipoprotein profiles, body composition, EE and RQ in obese and lean adult cats. We hypothesized that supplemental choline would result in greater serum lipid and lipoprotein concentrations, as compared to L-carnitine or control, whereas L-carnitine would increase EE and lower RQ only in the obese cats, suggesting favored fatty acid oxidation. Additionally, we hypothesized that both choline and L-carnitine would improve body composition by decreasing fat mass and increasing lean body mass only in the obese cats, as compared to control.

All procedures were approved by the University of Guelph Animal Care Committee (AUP#4496), in compliance with provincial and national guidelines regarding the care and use of animals in research.

One obese cat was removed during period three due to a medical condition not related to the study. As a result, data from this cat were not included for period three (choline treatment). Supplements were accepted by the cats with no observed adverse health effects throughout the trial.

To the authors' knowledge, this is the first study to investigate the effects of dietary choline supplementation in comparison to L-carnitine supplementation in obese and lean cats fed to maintain BW. Specifically, we sought to investigate lipotropic changes, as determined by serum lipid and lipoprotein profile, EE, RQ, and body composition data, between the two supplements. Previously, we reported that dietary choline at 5 and 6 x NRC RA increased serum lipid and lipoprotein profiles in overweight and obese cats, likely representing increased mobilization of hepatic lipids into circulation (21, 22). However, to allow the pet food industry and veterinary clinicians to make informed decisions regarding dietary supplementation and feline weight loss, it is important to assess the lipotropic effects of choline at the previously determined dose of 6 x NRC RA against L-carnitine (22), a supplement that has been frequently researched and used by the pet food industry (26–28). The dose of L-carnitine supplemented to the cats in the present trial was similarly based on previous research in overweight and obese cats (26, 28, 47, 48).

As expected, obese cats had greater food intake, energy intake, BW, BCS, and body composition values (total tissue mass, BF %, fat mass and LSTM), when compared to lean cats. These obese cats were specifically enrolled in this study due to their increased adiposity and were fed at energy intakes to support their current BW and BCS. A limitation of the present study was the approach to defining adiposity at enrollment, as the cats were classified as obese based solely on BCS. Body composition was not assessed by DXA at enrollment to evaluate BF % of the cats before group allocation and treatment. Although BF % differed between the lean and obese groups, the assigned mean BCS of the obese cats was higher than the respective mean BF % analyzed by DXA. The mean BF % of the obese cats was 32%, which corresponds with BCS 7/9 (36). Overall, all cats in the obese group could still be considered overweight or obese based on BF % (36), and BF % was still greater than in the lean cats. Because BCS is a subjective method of assessment, the BCS of the cats may have been overestimated due to localized fat deposits and/or abdominal fat pads which may have consisted of skin only. Similar limitations in investigator bias have previously been published in feline research (49, 50). Furthermore, differences in DXA machine, protocol, and computer software used by researchers can lead to differences in body composition and fat mass measurements among feline obesity studies (51).

Obesity results in alterations in the secretion of adipokines, leading to the development of other secondary health concerns such as insulin resistance and dyslipidemia (52–54). The obese cats in the present study had differences in their serum lipid and lipoprotein profiles, independent of treatment, when compared to the lean cats. Specifically, obese cats had overall concentrations of TAG and VLDL that were 59% greater than the lean cats. Additionally, overall concentrations of serum HDL-C were 13% lower in the obese cats compared to lean cats. To the authors' knowledge, there are no published reference ranges in cats for serum TAG, VLDL or HDL-C. However, the concentrations of serum TAG, VLDL and HDL-C reported herein were similar to those previously reported in overweight and obese adult cats receiving supplemental choline (21, 22). Serum TAG concentrations were also similar to those previously reported in healthy adult cats (55, 56). The present results align with previous reports of similarly higher circulating TAG, VLDL and lower HDL-C in obese cats, as compared to a lean control group (57–60). However, said authors also reported greater circulating NEFA, CHOL, and LDL-C concentrations in obese cats, which was not observed in the present study. As the mean age of the cats in the present study was lower than the mean age reported in these previous studies (57, 60), it is unclear if age may have also had an impact. To our knowledge, the effect of age on serum lipid, biochemistry and/or metabolomic profiles in overweight or obese cats has not specifically been investigated. However, cats under the age of two were reported to have lower serum CHOL and LDL-C concentrations as compared to cats over the age of two (BCS unknown) (61). Additionally, as these cats gained weight during a 12-week modified ad libitum feeding protocol during growth (25), there is a question of whether acute vs. chronic obesity may also impact lipoprotein profile and circulating metabolites. It is also worth considering that the increase in serum LDL and CHOL with choline treatment in both the lean and obese cats in the present study may have masked an effect of body condition that may have otherwise been observed. The same is possible for serum NEFA, where L-carnitine had a tendency to increase concentrations across both lean and obese cats.

The choline treatment in the present study provided a daily dietary choline intake of 6 x NRC RA, whereas cats on the L-carnitine and control treatments consumed a daily average choline intake of 1.2 and 1.3 x NRC RA, respectively. Both lean and obese cats had greater concentrations of CHOL, HDL-C-and LDL-C with the choline treatment, as compared to L-carnitine treatment (11, 7 and 33 % greater, respectively) and control (9, 9 and 13% greater, respectively). Serum CHOL concentrations remained within reference range for all treatments (2–12 mmol/L), and serum HDL-C and LDL-C concentrations were similar to those previously reported in obese and overweight adult cats consuming supplemental choline (21, 22). The greater concentrations of CHOL, HDL-C and LDL-C with choline in the present study are not surprising as similar changes were previously observed in obese and overweight cats consuming choline at 5 and 6 x NRC RA daily, respectively (21, 22). Choline is a precursor for PC, which is considered essential in the formation of VLDL. The VLDL are responsible for packaging TAG and CHOL, and exporting them out of the liver and into circulation (16). Similarly, both HDL-C and LCL-C have important roles in CHOL transport and circulation throughout the body, and require PC for their assembly (62, 63). Nascent HDL-C also requires PC for its formation within the liver (64). Contrary to what was observed, it was expected that serum TAG and VLDL would increase with choline supplementation in the present study. The present findings are contrary to our previous findings in obese and overweight cats consuming choline at 5 and 6 x NRC RA, where both serum TAG and VLDL concentrations were greater as compared to control (21, 22). Instead, obese cats receiving the L-carnitine treatment tended to have the greatest concentrations of serum TAG and VLDL. These results contradict the current research that exists on L-carnitine supplementation in hyperlipidemic animals and obese human patients with non-alcoholic steatohepatitis, in which TAG concentrations were reduced or unchanged (65–69). Although Rahbar et al. (70) observed increases in fasting plasma TAG concentrations when L-carnitine was supplemented to human patients with type II diabetes mellitus, the number of cats enrolled in the current study and the differences in BF % between the lean and obese group may not have been sufficient to detect these differences.

Although fasted serum glucose concentrations remained within reference range (4.4–7.7 mmol/L) for all the cats, glucose tended to be 16% greater in the obese cats than in the lean cats in the present study. Impaired glucose tolerance and insulin resistance are common findings in obese cats (9, 71–73), and a 30% decrease in insulin sensitivity has been reported with each additional kg of BW in cats (73). Over time, the decrease in insulin sensitivity may result in the development of type II diabetes mellitus (74), which is considered four times more likely in obese cats, as compared to lean cats (75). It is unclear if the increased TAG concentrations in the obese cats receiving L-carnitine may have been reflective of decreased glucose tolerance. Similar results have not been published in cats, and the use of L-carnitine in the treatment of obesity and diabetes mellitus in cats requires further investigation.

Serum ALP concentrations in the present study were within reference range for all cats throughout all three treatment periods (12–60 U/L), but tended to be, on average, 22% higher in the obese cats when compared to the lean cats. To the author's knowledge, there have been no published reports of increased ALP concentrations in healthy obese cats. However, in humans, a significant positive linear relationship between ALP and body mass index has been reported (76); where obese subjects had ALP concentrations that were on average 15 and 16% greater when compared to the non-obese and lean subjects, respectively (76, 77). This is because adipose tissue can express and release ALP into circulation (77). In the present study, it is likely that the greater serum ALP concentrations in the obese cats were representative of their increased adiposity, instead of their liver health. None of the cats showed any clinical signs of hepatic malfunction at any point throughout the study. However, serum liver enzyme values alone cannot adequately assess feline hepatic health and/or function. In the future, ultrasonography and histopathology of liver biopsies should be considered, along with pre- and postprandial bile acid profiles.

Conversely, lean cats consuming choline had the greatest concentrations of serum ALT. However, the concentrations similarly remained within reference range (31–105 U/L), suggesting that hepatobiliary diseases were absent in the present cats (78). Serum ALT concentrations were 26% greater in lean cats receiving the choline treatment as compared to control, and 20% greater as compared to the L-carnitine supplement. The present finding is unexpected, as ALT has been reported to increase in cases of non-alcoholic steatohepatitis resulting from choline deficiency in humans and in rodent models (79, 80). Additionally, these results contradict previous work by Verbrugghe et al. (81), where obese cats had higher ALT concentrations as compared to lean cats. Previously, choline supplementation did not affect serum ALT levels in overweight or obese cats (21, 22). The cats in the present study did not show signs of hepatic health concerns, regardless of treatment or body condition. However, as mentioned, hepatic health was not thoroughly investigated in the present study.

L-carnitine supplementation resulted in lower serum CREAT concentrations, as compared to the other two dietary treatments. This finding aligns with previous research in rodent models, where L-carnitine levels were inversely correlated with serum CREAT and renal function was improved with supplementation (82). The protective effects of L-carnitine on renal function are believed to be a result of its ability to reduce lipid peroxidation and free radical generation (83). Similarly, adult cats with chronic kidney disease (International Renal Interest Society Stages 1–2) consuming an average of 211 mg L-carnitine/kg BW from a test diet had serum CREAT levels that increased at a slower rate, as compared to the control group consuming an average of 2 mg L-carnitine/kg BW. However, the test diet and control diet differed in their overall nutrient profiles, and therefore the change in serum CREAT cannot be attributed to solely dietary L-carnitine intake (84). All cats in the present study were fed the same base diet, deemed in good health prior to enrollment with a complete blood count and serum biochemistry profile, and showed no signs of renal health concerns throughout the trial. Urinalysis and symmetric dimethyl arginine concentrations were not evaluated but should be considered for future L-carnitine studies.

Food and energy intakes were 0.3 % lower with choline treatment when compared to L-carnitine treatment, and 3% lower when compared to control. It is unclear whether the differences in food intake between treatments may have affected circulating metabolite concentrations, and to what effect. However, this reduction was not enough to influence BW, BCS, BF % and/or fat mass over the treatment periods. Similarly, it is also unlikely that this level of energy reduction was enough to put the cats in a negative energy balance and cause the observed lipid mobilization with choline treatment. It is unclear whether the change in food intake was due to palatability or due to a potential effect of satiety. A similar reduction in food intake was previously reported in growing kittens consuming choline at 3 x NRC RA. However, there were no changes in the analyzed fasted serum satiety hormone concentrations (ghrelin, leptin, glucagon-like peptide-1, peptide-YY or gastric inhibitory polypeptide) of said kittens with choline treatment (25). The mechanism through which choline may influence food intake in cats requires further elucidation.

Although there were no differences between dietary treatments on FM or BF %, there was a trend for greater LSTM in both obese and lean cats with treatment. On average, LSTM was 1.5% greater with choline treatment as compared to control, and 0.8% greater as compared to L-carnitine treatment. Supplementing betaine (a derivative of choline) in livestock species such as pigs has resulted in increased LSTM (23, 24). Although the exact mechanism has not been identified, it is presumably due to the role that betaine has as a possible methyl donor for the re-methylation of methionine; thus sparing methionine for protein synthesis and leading to increased s-adenosylmethionine production (85). There were no differences in LSTM with choline supplementation at 3 x NRC RA in growing kittens being fed ad libitum (25). However, supplemented kittens had lower food and subsequent protein intakes which may have impacted these results. In the present study, LSTM was 0.8% greater with L-carnitine compared to control. This observation aligns with previous research by Center et al. (26) in which overweight cats consumed a weight loss diet supplemented with 0 (control), 50, 100, or 150 mg carnitine/kg diet. Cats supplemented with carnitine achieved an increase in LSTM and a decrease in FM sooner (by day 42) compared to the cats receiving control (by day 84). These differences were attributed to lower RQ with treatment, indicating increased fat oxidation. Similarly, underweight and lean cats (BCS: 2.5–4/5) fed at 120% MER had lower fat deposition when supplemented 188 mg L-carnitine/kg diet compared to 121 mg L-carnitine/kg diet (28). However, the mechanism of action was unclear as there were no changes in EE, RQ or voluntary physical activity with dose throughout said 16-week trial.

Although EE is dependent on LSTM (86), the differences in LSTM between choline and the remaining treatments in the present study were minor and were likely not enough to elicit changes in EE. Instead, choline tended to increase RQ in the 120-min postprandial period, as compared to the other two treatments. The increase in RQ immediately post-feeding (0–120 min) may suggest a greater reliance on carbohydrate oxidation and less on gluconeogenesis with choline supplementation. This finding was not previously observed in overweight and obese cats or in guinea pigs supplemented with choline (21, 22, 87). It remains unclear why this may have occurred in the present study. There was a trend for greater fasted EE across all cats with the L-carnitine treatment. These findings align with previous research investigating L-carnitine supplementation in overweight cats. Overweight cats receiving 100 mg L-carnitine/kg BW similarly had greater fasted EE after 21 and 42 days of supplementation, as compared to overweight cats fed control (27). Although fed EE did not change with treatment in the present study, there was a trend for lower fed EE and fed RQ in the obese cats, as compared to the lean cats. Lower EE in overweight and obese cats as compared to lean cats is a common finding as an animal's surface area will influence EE (27, 86, 88). Perseghin et al. (89) similarly observed that overweight human participants had lower fed RQ when compared to lean participants, although not significant. The authors suggested that the increased leptin in the overweight participants may have played a role in promoting fatty acids toward oxidation instead of storage (90). However, said overweight study participants had normal insulin and leptin sensitivity. Although there are reports of decreased insulin and leptin sensitivity with weight gain in cats (71, 91), neither leptin nor insulin were measured in the obese cats in the present study and we therefore cannot conclude if the same association was present.

In agreement with previous publications, the present study found clear differences in circulating lipid and lipoprotein concentrations of obese and lean cats. Specifically, obese cats had greater concentrations of circulating TAG and VLDL, and lower concentrations of circulating HDL-C. Despite the differences associated with obesity, choline supplementation at six times the published RA resulted in greater HDL-C, CHOL and LDL-C in all cats. These findings are in alignment with our previous research in overweight and obese cats, which similarly found greater circulating lipid and lipoprotein concentrations with choline supplementation. These results suggest that choline supplementation may be beneficial in mobilizing lipids out of the liver and into circulation. Based on the findings of the present research, choline shows more promise than L-carnitine as a supplement for combating the risk of FHL in cats. While choline decreased food intake and had a tendency to improve LSTM, it did not increase EE in the same manner that L-carnitine did. Thus, future research should focus on evaluating the synergistic benefits of supplementing dietary choline and L-carnitine together to obese cats on a weight loss program. Together, L-carnitine and choline may prove useful in promoting weight loss, while maintaining LSTM and minimizing hepatic lipid stores in these cats.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was reviewed and approved by University of Guelph Animal Care Committee (AUP#4496).

AV, MB, and GK secured funding for this research. AR, AS, MB, GK, and AV were responsible for designing this study. AR, SV-S, and AV were responsible for data collection. AR was responsible for statistical analysis and writing of the manuscript. All authors contributed to editing and reviewing this manuscript.