Oral Presentation AUS-oMicS 2025

Investigating the Diabetic Lipidome: A DBS-Based High-Resolution Mass Spectrometry Approach (#43)

Jayden Lee Roberts 1 , Luke Whiley 1 , Nicola Gray 1 , Melvin Gay 2 , Elaine Holmes 1 3 , Julien Wist 1 4 , Jeremy K Nicholson 1 5 , Nathan G Lawler 1 3
  1. Australian National Phenome Centre, Murdoch, WA, Australia
  2. Bruker Daltonics, Preston, VIC
  3. Centre for Computational and Systems Medicine, Murdoch, WA, Australia
  4. Chemistry Department, Universidad del Valle, Cali, Colombia
  5. Institute of Global Health Innovation, Faculty of Medicine, Imperial College London, London, United Kingdom

 

Introduction: Comprehensive lipidomic analyses have identified distinct lipid signatures in diabetic patients, indicating that novel lipid markers may aid personalised approaches to monitoring disease. Dried blood spot (DBS) microsamples offer a minimally-invasive and cost-effective method for collecting and storing small sample volumes (< 50µL) for this purpose. Additionally, DBS enable increased sampling frequency and simplified transport via mail, suitable for remote/resource-limited settings. However, translation of lipidomic profiling of DBS for the diabetic context is limited, and further research is required before routine application of DBS can occur in longitudinal, decentralised, and remote settings.

 

Methods: Participants (n=17 diabetic, n=17 non-diabetic controls) performed DBS self-collection using advanced microsampling devices (10µL, Capitainer B®). Samples were analysed using an optimised untargeted 4D (timsTOF) method for semi-quantitative lipid analysis using ultrahigh-performance liquid chromatography and high-resolution mass spectrometry1. The 15-minute workflow enables analysis of 96 samples per day, covering 25 lipid subclasses spanning fatty-acyls, glycerolipids, glycerophospholipids, sphingolipids, and sterols (n = 432 unique lipids).

 

Results and Discussion: DBS lipid profiling revealed significant differences between diabetic individuals and non-diabetic controls (R²Y = 0.77, Q² = 0.41, p = 0.05). The lipid species driving this difference were consistent with those in matched traditional venous plasma samples (R²Y = 0.83, Q² = 0.49, p = 0.05) and capillary plasma samples (R²Y = 0.72, Q² = 0.49, p = 0.05), highlighting self-collected DBS as a reliable alternative to venous phlebotomy. Associations between lipid species and diabetes-related measures (Problem Areas in Diabetes scale - PAID-20, and Patient Activation Measure - PAM-13), indicate lipidomics may provide insight into relationships between lipid profiles, psychosocial factors, and health-management behaviors in diabetes.


Conclusion: Taken together with our previously demonstrated stability of DBS lipids1, preliminary findings indicate the feasibility for longitudinal and remote self-sampling with DBS, enabling personalised health monitoring for diabetes.

 

References:

679c8531195fa-Screenshot+2025-01-31+at+4.11.20%E2%80%AFpm.png