The current method for diagnosing methamphetamine use disorder (MUD) relies on self-reports and interviews with psychiatrists, which lack scientific rigor. This highlights the need for novel biomarkers to accurately diagnose MUD. In this study, we identified transcriptome biomarkers using hair follicles and proposed a diagnostic model for monitoring the MUD treatment process. We performed RNA sequencing analysis on hair follicle cells from healthy controls and former and current MUD patients who had been detained in the past for illegal use of methamphetamine (MA). We selected candidate genes for monitoring MUD patients by performing multivariate analysis methods, such as PCA and PLS-DA, and PPI network analysis. We developed a two-stage diagnostic model using multivariate ROC analysis based on the PLS-DA method. We constructed a two-step prediction model for MUD diagnosis using multivariate ROC analysis, including 10 biomarkers. The first step model, which distinguishes non-recovered patients from others, showed very high accuracy (prediction accuracy, 98.7%). The second step model, which distinguishes almost-recovered patients from healthy controls, showed high accuracy (prediction accuracy, 81.3%). This study is the first report to use hair follicles of MUD patients and to develop a MUD prediction model based on transcriptomic biomarkers, which offers a potential solution to improve the accuracy of MUD diagnosis and may lead to the development of better pharmacological treatments for the disorder in the future.
Keywords: RNA sequencing; methamphetamine; methamphetamine use disorder; peripheral biomarker; prediction model.