Novel urinary metabolite signature for diagnosing postpartum depression.

Abstract

BackgroundPostpartum depression (PPD) could affect ~10% of women and impair the quality of mother–infant interactions. Currently, there are no objective methods to diagnose PPD. Therefore, this study was conducted to identify potential biomarkers for diagnosing PPD.Materials and methodsMorning urine samples of PPD subjects, postpartum women without depression (PPWD) and healthy controls (HCs) were collected. The gas chromatography-mass spectroscopy (GC-MS)-based urinary metabolomic approach was performed to characterize the urinary metabolic profiling. The orthogonal partial least-squares-discriminant analysis (OPLS-DA) was used to identify the differential metabolites. The logistic regression analysis and Bayesian information criterion rule were further used to identify the potential biomarker panel. The receiver operating characteristic curve analysis was conducted to evaluate the diagnostic performance of the identified potential biomarker panel.ResultsTotally, 73 PPD subjects, 73 PPWD and 74 HCs were included, and 68 metabolites were identified using GC-MS. The OPLS-DA model showed that there were 22 differential metabolites (14 upregulated and 8 downregulated) responsible for separating PPD subjects from HCs and PPWD. Meanwhile, a panel of five potential biomarkers – formate, succinate, 1-methylhistidine, α-glucose and dimethylamine – was identified. This panel could effectively distinguish PPD subjects from HCs and PPWD with an area under the curve (AUC) curve of 0.948 in the training set and 0.944 in the testing set.ConclusionThese results demonstrated that the potential biomarker panel could aid in the future development of an objective diagnostic method for PPD.