New test that measures very small molecules in the blood assesses the risk of autoimmune diseases using artificial intelligence
Analysis of small molecules in the blood (metabolites), in combination with lifestyle factors, such as exercise and alcohol, can predict the likelihood of having an autoimmune disease by 93%. The study was published in Frontiers in Molecular Biosciences - Metabolomics in October.
The researchers used targeted metabolomics and compared the metabolic profile of patients with autoimmune disease with the metabolic profile of healthy individuals, using three different statistical models and artificial neural network.
Autoimmune Diseases & Modern Lifestyle
Autoimmune diseases are increasing rapidly worldwide. There is growing evidence suggesting that the underlying cause of autoimmune diseases is the metabolic disruption that is directly linked to the modern lifestyle and diet. However, these factors are not adequately addressed in the current medical approach.
This is the first test that integrates the metabolomic profile and lifestyle habits of an individual to estimate the risk of Autoimmune Diseases
This new test brings metabolomics closer to clinical application as an adjunct tool for doctors addressing autoimmune diseases. The researchers that participated in the study developed three different prediction models to estimate the likelihood of autoimmune disease presence reaching a predictive accuracy of 93%.
The input consists of blood analysis of total fatty acids quantified by Gas Chromatography/ Mass Spectrometry and the modifiable risk factors of B.M.I., gender, age, exercise, and alcohol consumption. Overall, 28 variables were analyzed in order to predict the risk of several autoimmune diseases, including Hashimoto's thyroiditis, rheumatoid arthritis, psoriasis, multiple sclerosis, inflammatory bowel disease, vitiligo and other less frequent autoimmune diseases.
The model was able to predict correctly up to 93% of those with ADs and 78.9% for the total population in a retrospective nested case control study of 403 individuals, while important biomarkers were identified through artificial neural networks (ANN).
These findings suggest that a particular metabolic profile is associated with the presence of autoimmune diseases and can be used as an additional tool in the early diagnosis of autoimmune diseases.
Metabolomics is increasingly used in the research aiming to discover and identify biomarkers for chronic diseases. A separate metabolic fingerprint of autoimmune diseases can help prevent and diagnose early. "Most patients with autoimmune diseases are diagnosed months, or years after the first symptoms of the disease, making treatment more difficult. The predictive model is the result of years of research and will greatly help in the early detection and treatment, " says Dr. Dimitris Tsoukalas , head of the study.
The prediction model was developed by the Metabolomic Medicine ® research team in collaboration with researchers from International and Greek Universities, including Albert Einstein School of Medicine in New York, USA and the Golden Helix Foundation in Oxford, United Kingdom.
Targeted Metabolomic Analysis of Serum Fatty Acids for the Prediction of Autoimmune Diseases
Dimitris Tsoukalas1,2,3, Vassileios Fragoulakis4, Evangelia Sarandi2,5, Anca Oana Docea6, Evangelos Papakonstaninou2, Gerasimos Tsilimidos2, Chrysanthi Anamaterou2, Persefoni Fragkiadaki5, Michael Aschner7, Aristidis Tsatsakis3,5, Nikolaos Drakoulis8 and Daniela Calina1
ORIGINAL RESEARCH ARTICLE
Front. Mol. Biosci., 01 November 2019 | https://doi.org/10.3389/fmolb.2019.00120
1Department of Clinical Pharmacy, Faculty of Pharmacy, University of Medicine and Pharmacy, Craiova, Romania
2Metabolomic Medicine, Health Clinic for Autoimmune and Chronic Diseases, Athens, Greece
3E.INu.M, European Institute of Nutritional Medicine, Rome, Italy
4The Golden Helix Foundation, London, United Kingdom
5Laboratory of Toxicology and Forensic Sciences, Medical School, University of Crete, Heraklion, Greece
6Department of Toxicology, Faculty of Pharmacy, University of Medicine and Pharmacy, Craiova, Romania
7Department of Molecular Pharmacology, Albert Einstein College of Medicine, The Bronx, NY, United States
8Research Group of Clinical Pharmacology and Pharmacogenomics, Faculty of Pharmacy, School of Health Sciences, National and Kapodistrian University of Athens, Athens, Greece