Litcius/Paper detail

Comparing immunoassay and mass spectrometry techniques for salivary sex hormone analysis

Alexandra Brouillard, Lisa‐Marie Davignon, Rebecca Cernik, Charles‐Édouard Giguère, Helen Findlay, Robert‐Paul Juster, Sonia Lupien, Marie‐France Marin

2025Psychoneuroendocrinology18 citationsDOIOpen Access PDF

Abstract

From a confluence of events, our team acquired salivary sex hormone data from two different assays; namely, enzyme-linked immunosorbent immunoassay (ELISA; Salimetrics) and liquid chromatography-tandem mass spectrometry (LC-MS/MS). As previous research has often discussed inter-assay differences but lacked direct comparative data for these specific hormones in saliva, this paper compared both techniques on their ability to accurately quantify concentrations of estradiol, progesterone, and testosterone in healthy young adults (72 combined oral contraceptive [COC] users, 99 naturally cycling [NC] women in the early follicular and pre-ovulatory phases, and 47 men). Using multivariate and computational approaches, our results converged and showed poor performance of ELISA for measuring salivary sex hormones, with estradiol and progesterone being much less valid than testosterone. Despite its challenges with quantification, LC-MS/MS was found to be superior. Our study underscores the importance of methodological rigor in sex steroid hormone assay techniques, highlighting LC-MS/MS as a more reliable option compared to ELISA for salivary sex hormone quantification in healthy adults. These findings contribute to the ongoing dialogue in the field concerning the validity and reproducibility of scientific discoveries. Indeed, accurate measurement is crucial for generating reliable findings regarding the intricate relationships between hormones, brain, behavior, and mental health. • We assessed ELISA and LC-MS/MS performance for measuring salivary sex hormones. • The between-methods relationship was strong for salivary testosterone only. • LC-MS/MS showed expected differences in estradiol and testosterone in women. • Machine-learning classification models revealed better results with LC-MS/MS. • LC-MS/MS promises to improve validity of sex steroid profiling of healthy adults.

Topics & Concepts

ImmunoassayMass spectrometryChromatographySex hormone-binding globulinChemistryHormoneMedicineInternal medicineAntibodyImmunologyAndrogenStress Responses and CortisolSalivary Gland Disorders and FunctionsHormonal and reproductive studies