Biomarkers of non-communicable chronic disease: an update on contemporary methods

Abstract

Chronic diseases constitute a major global burden with significant impact on health systems, economies, and quality of life. Chronic diseases include a broad range of diseases that can be communicable or non-communicable. Chronic diseases are often associated with modifications of normal physiological levels of various analytes that are routinely measured in serum and other body fluids, as well as pathological findings, such as chronic inflammation, oxidative stress, and mitochondrial dysfunction. Identification of at-risk populations, early diagnosis, and prediction of prognosis play a major role in preventing or reducing the burden of chronic diseases. Biomarkers are tools that are used by health professionals to aid in the identification and management of chronic diseases. Biomarkers can be diagnostic, predictive, or prognostic. Several individual or grouped biomarkers have been used successfully in the diagnosis and prediction of certain chronic diseases, however, it is generally accepted that a more sophisticated approach to link and interpret various biomarkers involved in chronic disease is necessary to improve our current procedures. In order to ensure a comprehensive and unbiased coverage of the literature, first a primary frame of the manuscript (title, headings and subheadings) was drafted by the authors working on this paper. Second, based on the components drafted in the preliminary skeleton a comprehensive search of the literature was performed using the PubMed and Google Scholar search engines. Multiple keywords related to the topic were used. Out of screened papers, only 190 papers, which are the most relevant, and recent articles were selected to cover the topic in relation to etiological mechanisms of different chronic diseases, the most recently used biomarkers of chronic diseases and finally the advances in the applications of multivariate biomarkers of chronic diseases as statistical and clinically applied tool for the early diagnosis of chronic diseases was discussed. Recently, multivariate biomarkers analysis approach has been employed with promising prospect. A brief discussion of the multivariate approach for the early diagnosis of the most common chronic diseases was highlighted in this review. The use of diagnostic algorithms might show the way for novel criteria and enhanced diagnostic effectiveness inpatients with one or numerous non-communicable chronic diseases. The search for new relevant biomarkers for the better diagnosis of patients with non-communicable chronic diseases according to the risk of progression, sickness, and fatality is ongoing. It is important to determine whether the newly identified biomarkers are purely associations or real biomarkers of underlying pathophysiological processes. Use of multivariate analysis could be of great importance in this regard.

Keywords: Biomarkers; Chronic diseases; Inflammation; Mitochondrial dysfunction; Multivariate analysis; Oxidative stress; Principal component analysis.

Figure 1. Risk factors of non-communicable chronic diseases.

Figure 2. Non-communicable and communicable of chronic diseases.

Figure 3. Uncoupling proteins as mitochondrial dysfunction-related etiological mechanisms of most chronic diseases.

Figure 4. Simplified illustration of D-dimer generation as late-stage biomarker of cardiovascular diseases.

Thrombin transforms fibrinogen to fibrin monomers that are composed of a central E-domain and 2 outer D-domains. Fibrin monomers polymerize, thus making an unstable fibrin mesh. Activated blood clotting factor XIII cross-links the D-domains, which supports the fibrin network. Fibrin-bound plasmin degrades the fibrin mesh into soluble fragments: D-dimers and E fragments.

Figure 5. Comparing principal component analysis and discriminant analysis.

Principal component analysis (PCA) and discriminant analysis (DA) both combine correlated variables into common vectors (A–D) known as principal components (PCs) and discriminant functions (DSCs) in PCA and DA, respectively. The first component (PC1) in principal component analysis is designed to account for the most variance, while ignoring predefined groups (A & B). The first discriminant function (DSC1) in discriminant analysis is designed to maximize separation between predefined group (C & D). PCs and DSCs can be rotated and used to plot the data in a new coordinate system (B & D). Data variation (variance) aligned with PC1 and DSC1 are illustrated (E). Dotted axes (B & D) represent the original directions of PCs and DSCs before rotation.