Metabolomics - High resolution characterization of human metabolism and biochemical status on habitual diet and ketogenic diet

Introduction and background: Ketogenic diet (KD) consisting of mostly fat, adequate protein and very low carbohydrate consumption induces metabolic modifications where the main fuel switches from glucose to fat degradation. The diet is used in the treatment of several inborn errors of metabolism and as treatment for refractory epilepsy. The diet has also become popular as a rapid weight reduction method. High intake of fat, particularly saturated fat, can lead to high levels of total and LDL cholesterol concentrations known to be associated with development of atherosclerosis. This raises a concern that KD increases the risk of cardiovascular disease. However, the biochemical effects of KD are insufficiently described, and previous studies have been conducted primarily using overweight participants. Objectives: This study aimed to identify metabolic changes in normal weight subjects as a result of following a ketogenic diet for three weeks. In addition, comparison between the ketogenic and habitual diet metabolomes of subjects with low and high relative LDL-C change after KD intervention was conducted. Subjects and methods: In 2011-2012, a six-weeks, randomized controlled cross-over trial investigating the effect of three weeks on ketogenic diet (<20 g carbohydrates) was conducted in healthy normal weight subjects (N=30), mainly students and employees at the Department of Nutrition research at University of Oslo. As a continuation, comprehensive global metabolomics analyses of plasma with high-performance liquid chromatography coupled to mass spectrometry (LC-MS) was performed in this thesis to identify and quantify as many metabolites as possible in the biological samples. Results: An average of 1085 and 419 features were detected in positive and negative ionization mode, respectively. Differential analysis using volcano plot showed that 48 features in positive ionization mode and 32 compounds in negative ionization mode were detected in significantly higher levels in samples taken after KD versus habitual diet with a log2 fold change >1. Significantly higher levels of ketone bodies, lipids, acylcarnitines, trihomomethionine, acetylglycine, GABA and hydroxyvaleric were detected. Hydroxyvaleric acid had a greater increase in the group with the highest relative LDL-C increase compared to the lowest relative LDL-C increase group. Furthermore, citric acid, malic acid and alpha-ketoglutaric acid as intermediated in the citric acid cycle were significantly upregulated after KD. Levels of the amino acids threonine, asparagine, tryptophan, proline, b-alanine, leucine, arginine, methionine, glutamine and alanine were significantly decreased, while valine was elevated 43 % after 3 weeks on KD. Univariate correlation matrix showed a significant positive correlation (p<0.03) between the amount of amino acid histidine and change in LDL-C with a correlation coefficient of 0.44 (CI 0.06, 0.71). Lower plasma levels of stachydrine, methyl indole 3-acetate, caffeine and N-acetylvaline was observed after KD. Conclusion: Applying global metabolomics in a randomized controlled study has the potential to detect differences in the metabolome after dietary intervention, and further increase our knowledge of the biochemical effects and consequences of a KD. In particular, we found alterations in a diverse set of compounds. However, further investigations are needed to understand the mechanisms behind these changes and to determine long-term effects of following a KD.