Salivary Prevotella qPCR Signal as an Exploratory Non-Invasive Adjunct for Rotterdam Phenotype Stratification in Women with Polycystic Ovary Syndrome: A Proof-of-Concept Cross-Sectional Study

Background/Objectives: Polycystic ovary syndrome (PCOS) is a heterogeneous endocrine-metabolic disorder. Rotterdam-defined phenotypes may need adjunctive, non-invasive signals for research-level phenotypic stratification. Saliva deserves attention because it is easy to collect repeatedly, acceptable in outpatient settings, and may reflect oral-systemic inflammatory and endocrine-metabolic interactions. This study evaluated whether salivary qPCR signals for Lactobacillus, Prevotella, and Bifidobacterium differ across Rotterdam-defined PCOS phenotypes and controls, with emphasis on the exploratory relevance of Prevotella. Methods: This cross-sectional proof-of-concept study included 110 women: 87 with PCOS and 23 controls. PCOS phenotypes were classified according to the Rotterdam criteria. Salivary microbial targets were assessed using SYBR Green-based genus-specific qPCR. Available instrument-export spreadsheets were reviewed for standard-curve quality control. Because the available assay outputs did not support robust absolute quantification, inferential analyses used Cq-based microbial signals only. Lower Cq values indicate stronger target DNA signal. No universal bacterial 16S rRNA reference, exogenous spike-in, salivary flow correction, or fully validated copy-number conversion was available. Group differences were evaluated using non-parametric tests with Bonferroni-corrected post-hoc comparisons. ROC, regression, and correlation analyses were retained as hypothesis-generating analyses only. Results: Prevotella Cq values differed significantly across groups (p < 0.001), with lower median Cq values in phenotypes A, B, and C than in phenotype D and controls. Lactobacillus Cq values did not differ significantly (p = 0.249). Bifidobacterium showed an overall group difference (p < 0.001), but its pattern and assay performance were less consistent. Among women with PCOS, Prevotella Cq values were associated with Ferriman-Gallwey score and polycystic ovarian morphology. Conclusions: Salivary Prevotella showed the clearest exploratory genus-level qPCR signal across Rotterdam-defined PCOS phenotypes. The findings support further technical and clinical validation of saliva-based microbial profiling as a possible adjunct to conventional PCOS phenotyping. They do not validate Prevotella as a standalone diagnostic biomarker, do not define clinical cutoffs, and do not quantify absolute bacterial load.