Dietary protein insufficiency has been linked to excessive TAG storage and non-alcoholic fatty liver disease (NAFLD) in developing countries. Hepatic TAG accumulation following a low-protein diet may be due to altered peroxisomal, mitochondrial and gut microbiota function. Hepatic peroxisomes and mitochondria normally mediate metabolism of nutrients to provide energy and substrates for lipogenesis. Peroxisome biogenesis and activities can be modulated by odd-chain fatty acids (OCFA) and SCFA that are derived from gut bacteria, for example, propionate and butyrate. Also produced during amino acid metabolism by peroxisomes and mitochondria, propionate and butyrate concentrations correlate inversely with risk of obesity, insulin resistance and NAFLD. In this horizon-scanning review, we have compiled available evidence on the effects of protein malnutrition on OCFA production, arising from loss in mitochondrial, peroxisomal and gut microbiota function, and its association with lipid accumulation in the liver. The methyl donor amino acid composition of dietary protein is an important contributor to liver function and lipid storage; the presence and abundance of dietary branched-chain amino acids can modulate the composition and metabolic activity of the gut microbiome and, on the other hand, can affect protective OCFA and SCFA production in the liver. In preclinical animal models fed with low-protein diets, specific amino acid supplementation can ameliorate fatty liver disease. The association between low dietary protein intake and fatty liver disease is underexplored and merits further investigation, particularly in vulnerable groups with dietary protein restriction in developing countries.

Consumption of diets containing medium-chain TAG (MCT) has been shown to confer neuroprotective effects. We aim to identify the global metabolic perturbations associated with consumption of a ketogenic diet (medium-chain TAG diet (MCTD)) in dogs with idiopathic epilepsy. We used ultra-performance liquid chromatography-MS (UPLC-MS) to generate metabolic and lipidomic profiles of fasted canine serum and made comparisons between the MCTD and standardised placebo diet phases. We identified metabolites that differed significantly between diet phases using metabolite fragmentation profiles generated by tandem MS (UPLC–MS/MS). Consumption of the MCTD resulted in significant differences in serum metabolic profiles when compared with the placebo diet, where sixteen altered lipid metabolites were identified. Consumption of the MCTD resulted in reduced abundances of palmitoylcarnitine, octadecenoylcarnitine, stearoylcarnitine and significant changes, both reduced and increased abundances, of phosphatidylcholine (PC) metabolites. There was a significant increase in abundance of the saturated C17 : 0 fatty acyl moieties during the MCTD phase. Lysophosphatidylcholine (17 : 0) (P=0·01) and PC (17:0/20:4) (P=0·03) were both significantly higher in abundance during the MCTD. The data presented in this study highlight global changes in lipid metabolism, and, of particular interest, in the C17 : 0 moieties, as a result of MCT consumption. Elucidating the global metabolic response of MCT consumption will not only improve the administration of current ketogenic diets for neurological disease models but also provides new avenues for research to develop better diet therapies with improved neuroprotective efficacies. Future studies should clarify the involvement and importance of C17 : 0 moieties in endogenous MCT metabolic pathways.