Metabolomics analysis of human serum for characterisation of phenotypically ageing men in different ethnic populations

  • Dakshat Trivedi

Student thesis: Phd


With increasing life expectancy in most developed countries, the promotion of healthy ageing is a significant public health objective. Changes in the human metabolome during this phase of life are relatively under-studied; by way of example, relatively little is known about the hormonal transitions in men compared to female at the onset and progression of ageing. Metabolomics is aimed at measuring the totality of the metabolites within a biological system, be it blood, urine, or cells. To achieve this gas chromatography (GC) and liquid chromatography (LC) separation coupled to mass spectrometry (MS) detection techniques will be used in these studies to analyse the serum metabolome. In addition to MS-based metabolomics, studies reported in this thesis also incorporate various analytical platforms for the analysis of these serum samples using vibrational spectroscopic techniques including as Raman spectroscopy and Fourier transform infrared (FT-IR) spectroscopy. This PhD project was aimed at analysing serum samples of men from a different ethnic background, to generate metabolic data, and then to link these metabolomes with ethnicity and age to study their influence on the metabolome. The FT-IR spectroscopy applied for metabolic fingerprinting poorly classified the cohort based on their age. However, there was significant classification seen between young men (40-49.99 years) from their older counterparts (≥ 60 years). The age models poorly classified those of age 50-59.99 years. Raman spectroscopy reported similar results and was unable to classify the cohort based on their age. However, both techniques were robust and accurate in classifying the cohort based on their ethnic origin. From the spectral signatures obtained from FT-IR spectroscopy, lipids, fatty acids and nucleic acids were of primary interest and significant contributors to classifications. Raman analysis too showed similar spectral information to corroborate the findings. MS-based methods were both able to classify subjects based on their age when split into groups of those younger than 60 years and those 60 years or above. A rigid grouping of the cohort into four groups of 10 years span (between 40- > 69.99) did not return a statistically significant classification. Ethnicity based classification models performed exceptionally well for both LC-MS and GC-MS analysis. LC-MS highlighted the TCA cycle (for age) and NGlycan biosynthesis pathway (for ethnicity) as being impacted by the metabolites contributing significantly to the classification. As for GC-MS, similar findings for perturbation in the TCA cycle associate to age group classification. Whilst for ethnicity-based classification, the fatty acid biosynthesis pathway was impacted. All metabolites of interest have been highlighted and discussed in the relevant sections of the chapters. This thesis, therefore, reports profiling results from a combination of four techniques of varying analytical strength and coverage of serum metabolome.
Date of Award1 Aug 2021
Original languageEnglish
Awarding Institution
  • The University of Manchester
SupervisorFrederick Wu (Supervisor), Perdita Barran (Supervisor) & Drupad Trivedi (Supervisor)


  • Metabolomics
  • Human Serum
  • Ageing
  • Ethnicity
  • Metabolic Profiling

Cite this