Diagnosis and treatment of widespread skin diseases reaches a new level
2–3% of the population and about 4% of adults suffer from psoriasis and atopic dermatitis. Another widespread disease is atherosclerosis, which affects the life quality of many older people. For the first time, machine-learning models developed at the University of Tartu Faculty of Medicine found markers that have never been associated with these diseases before. New knowledge helps diseases to be diagnosed better and to monitor the course of treatment. One day, when this model becomes available for physicians, diagnosing and treating widespread diseases will become more precise and cheaper.
According to Aigar Ottas, a doctoral student at the University of Tartu Institute of Biomedicine and Translational Medicine and a specialist of the Mass Spectrometry Laboratory, psoriasis and atopical dermatitis have many concomitant diseases. Atherosclerosis is one of these. Its risk factors are, for example, advanced age, male gender, smoking, obesity, alcohol consumption and high blood levels of LDL or the so-called bad cholesterol. “It is certain that psoriasis and atopic dermatitis contribute to the manifestation of atherosclerosis and that the health behaviour decisions of an individual affect their risk for developing various diseases.”
In his doctoral thesis, “The Metabolomic Profiling of Psoriasis, Atopic Dermatitis and Atherosclerosis”, Ottas characterised the three most common diseases – psoriasis, atopic dermatitis and atherosclerosis using various scientific methods and statistical modelling.
His research belongs to the field of metabolomics, which concerns the measurement and analysis of low molecular-weight compounds. Using various methods, it is possible to create a metabolomic fingerprint, which characterises certain diseases or conditions. In his thesis, Ottas was able to find, among other things, disease-related markers, which have never been associated with these diseases before. Thus, he discovered several products of metabolic pathways that are specific to the diseases.
For example, in the case of psoriasis, Ottas discovered multiple amino acids that can be associated with the rapid proliferation of skin cells and biomarkers that indicate oxidative stress. In the study of atopic dermatitis, he discovered many acylcarnitine ratios that are indicative of aberrations in fat metabolism and an increase in the risk of cardiovascular diseases. When analysing atherosclerosis, Ottas found phosphatidylcholines, which were directly correlated to arterial stiffness and an increased heart rate. In addition, multiple markers were found in psoriasis and atopic dermatitis that might increase the risk for cardiovascular diseases.
Based on the data gathered in the research, statistical machine-learning models were created, which can predict whether a sample belongs to a case or control with high probability.
“My thesis is preliminary work for more extensive research, which focuses on creating machine-learning models for predicting whether samples belong to a case or control. This will help physicians diagnose diseases better and monitor the course of treatment,” described Ottas and added that more accurate and cost-effective diagnosis and treatment of these diseases would greatly benefit the society. “When these solutions reach into everyday medicine, an individual suffering from psoriasis, atopic dermatitis or atherosclerosis is sure to get better treatment.”
Aigar Ottas defended his doctoral thesis “The Metabolomic Profiling of Psoriasis, Atopic Dermatitis and Atherosclerosis” on 18 June at Biomedicum, Tartu, Estonia.
Thesis supervisors are Ursel Soomets, a professor of medical metabolomics at University of Tartu Institute of Biomedicine and Translational Medicine, and Külli Kingo, a professor of dermatology and venerology at the Institute of Clinical Medicine.
The translation of this article was funded by the European Regional Development Fund through Estonian Research Council.