Infrared Thermography of Acupoints Shows Correlation with Depression Severity in Adolescents

BackgroundThe prevalence of adolescent major depressive disorder (MDD) is rising; however, diagnosis relies on subjective measures due to a lack of objective biomarkers. This study explored infrared thermography (IRT) as a non-invasive tool to quantify thermal radiation characteristics of acupoints in adolescents with MDD. The objective was to establish diagnostic models based on acupoint temperature-derived biomarkers.MethodsA prospective, multi-center observational study enrolled 108 participants (65 adolescents with MDD and 43 healthy controls [HCs]). We first examined correlations between acupoint temperatures and depression severity using Pearson analysis. Multiple linear and binary logistic regression models were developed to diagnose MDD and assess severity. The diagnostic model for MDD was visualized as a nomogram and validated using Receiver Operating Characteristic (ROC) curves, Hosmer-Lemeshow tests, calibration plots, and decision curve analysis (DCA). Internal validation was performed using the bootstrap method.ResultsAmong 27 acupoints analyzed, adolescents with MDD exhibited altered acupoint temperatures at Taiyang (EX-HN5), Quchi (LI11), Yanggu (SI5), and Waiqiu (GB36). Subsequent Pearson correlation analysis revealed negative correlations between the infrared relative temperatures of Taiyang (EX-HN5), Quchi (LI11), and Waiqiu (GB36) and depression severity (P = 0.001, r = -0.319; P = 0.022, r = -0.229; P = 0.001, r = -0.325) and a weak positive correlation between the infrared relative temperature of Yanggu (SI5) and depression severity (P = 0.043, r = 0.202). Building on these findings, two diagnostic models were developed: a linear regression model for depression severity of adolescents (Y = 52.25-9.52*TEX-HN5-13.07*TGB36) and a logistic regression model for adolescents with MDD diagnosis (P = ex/(1+ex), x = 0.22-1.14*TEX-HN5+0.45*TSI5-2.19*TGB36). The nomogram-based model demonstrated good calibration (Hosmer-Lemeshow P = 0.855), discrimination (AUC = 0.785, 95%CI: 0.693 - 0.876), and clinical utility. Internal validation using the bootstrap method produced a C-index of 0.752 (95% CI: 0.617 - 0.877), further confirming the model’s robustness.ConclusionsIn conclusion, acupoint temperature-based models show promising efficacy for the objective and non-invasive diagnosis and severity quantification of adolescents with MDD, offering valuable tools for early clinical intervention. Future studies should validate these findings across diverse populations and integrate multi-modal biomarkers to enhance diagnostic precision.Clinical Trial RegistrationClinicalTrials.gov, identifier NCT06750640.