The following is the summary of “Plasma Metabolomics Reveals Systemic Metabolic Alterations of Subclinical and Clinical Hypothyroidism” published in the January 2023 issue of Clinical Endocrinology & Metabolism by Shao, et al.
However, the underlying metabolic changes in clinical hypothyroidism and subclinical hypothyroidism (CH and SCH, respectively) remain unknown. To better understand the pathophysiology of hypothyroidism, metabolomics may be able to shed light on the metabolic processes involved. To further distinguish SCH and CH patients from euthyroid controls, researchers investigated metabolic changes in these conditions and identified candidate metabolite biomarkers. Metabolomics using high-resolution mass spectrometry was applied to plasma samples from 126 human participants, comprising 45 patients with CH, 41 patients with SCH, and 40 euthyroid controls.
Multivariate principal components analysis and orthogonal partial least squares discriminant analysis were used to process the data. Multivariate linear regression was used to examine the correlations. Models for making predictions based on possible metabolite biomarkers were developed using the unbiased Variable selection in R technique and 3 machine learning models. Significant differences were seen between the plasma metabolomic profiles of the SCH and CH groups and those of the control groups, yet, there were striking similarities in n the metabolite changes between the two groups. Primary bile acid production, steroid hormone biosynthesis, lysine degradation, tryptophan metabolism, and purine metabolism were all found to be significantly influenced by SCH and CH in a pathway enrichment analysis.
There were significant correlations between 65 metabolites and thyroid hormones like thyrotropin, free thyroxine, thyroid peroxidase antibody, and thyroglobulin antibody. To distinguish between the 3 groups, investigators opted for and validated 17 metabolic indicators. Because of the abnormal metabolic patterns associated with hypothyroidism in SCH and CH, metabolomics combined with machine learning algorithms can be utilized to create diagnostic models based on particular metabolites.