J Autoimmun. 2026 Apr 02. pii: S0896-8411(26)00030-2. [Epub ahead of print]160
103552
Anti-Müllerian hormone (AMH) is an early functional marker of diminished ovarian reserve and premature ovarian insufficiency (POI), yet the immunologic context underlying AMH variation remains incompletely defined. Because ovarian autoimmunity constitutes a distinct subset of POI, anti-ovarian antibody (AOA) status may modulate the immune-endocrine-metabolic mechanisms that govern follicular function. In this retrospective observational cohort study, 803 women evaluated for reproductive failure between 2018 and 2024 were stratified by AOA status into AOA-positive (n = 135) and AOA-negative (n = 668) groups. Serum AMH was assessed together with endocrine and metabolic parameters, including 25-hydroxy vitamin D (VitD), thyroid-stimulating hormone (TSH), dehydroepiandrosterone sulfate (DHEAS), plasminogen activator inhibitor-1 (PAI-1), homocysteine, HbA1c, and androgen indices, as well as peripheral immune markers comprising lymphocyte subsets, natural killer (NK) cell cytotoxicity, and T helper (Th) 1/Th2 cytokine producing Th cell ratios. Age-adjusted AMH ratios (aaAMH ratios) were calculated using a validated polynomial model based on U.S. reference data. Multivariable regression and generalized additive models (GAM) were applied to evaluate both linear and nonlinear associations between biological variables and ovarian reserve. Age was inversely associated with AMH and excluded from adjusted models. In AOA-negative women, aaAMH ratios demonstrated significant nonlinear associations with NK cell proportion and B-1a cells, suggesting preserved but complex immune regulatory dynamics. Conversely, in AOA-positive women, several markers showed significant linear associations with aaAMH ratios, including homocysteine (Hcy), PAI-1, HbA1c, and the IFN-γ/IL-10 Th1/Th2 cytokine-producing Th cell ratio. The GAM explained substantially greater deviance in AOA-positive women than in AOA-negative women, indicating a distinct systemic regulatory pattern. A composite model integrating metabolic and immune markers yielded an area under the curve of 0.786 for identifying women at risk of markedly reduced ovarian reserve. These findings reveal fundamentally different immune-endocrine-metabolic signatures of ovarian dysfunction according to ovarian autoimmunity and support the concept that peripheral immune profiles reflect immune-mediated modulation of follicular function in autoimmune POI.
Keywords: Anti-müllerian hormone; Anti-ovarian antibody; Immune-endocrine interaction; Natural killer cells; Ovarian autoimmunity; Premature ovarian insufficiency