Data mining and clinical observational studies to examine the link between smoking and premature ovarian insufficiency

Objective This study explores the link between smoking and premature ovarian insufficiency to inform smoking cessation strategies and protect women’s reproductive health. Methods Using data mining for literature retrieval and association rule analysis, we built a protein interaction network and identified core targets. GO and KEGG analyses revealed biological functions and pathways linked to early-onset ovarian dysfunction. We then measured urinary uterine cantrinin levels in Chinese patients with infrequent menstruation (May 2023–May 2025), grouped them by presence or absence of premature ovarian insufficiency, and statistically tested its association with the condition. Results Data mining identified six cigarette smoke components directly linked to impaired ovarian function: nicotine, benzo[a]pyrene (BaP), Nicotine-derived nitrosamine ketone (NNK), acrolein, 1,3-butadiene, and cotinine. Topological analysis confirmed their central role in tobacco-induced early-onset ovarian dysfunction. GO analysis showed they disrupt granulosa cell proliferation by interfering with the cell cycle; KEGG analysis linked them to progesterone-mediated oocyte maturation, cell cycle, and oocyte meiosis pathways. In a clinical study, higher urinary cotinine levels correlated with increased disease incidence (quartile analysis) and nonlinearly elevated recurrence risk (restricted cubic spline), especially above the median. Cotinine showed weak positive associations with Follicle stimulating hormone(FSH) and Luteinizing hormone(LH) but no significant link with Anti-Müllerian hormone(AMH). Logistic regression indicated moderate predictive value for premature ovarian insufficiency, persisting after adjustment and subgroup analysis. Notably, cotinine associated nonlinearly with AMH—minimal change below median, sharp rise above it—while maintaining linear associations with FSH and LH. Multivariable linear regression confirmed consistent positive correlations between cotinine and all three markers. Conclusion Smoking is linked to premature ovarian insufficiency through a well-defined molecular mechanism, and urinary cotinine levels correlate significantly with both disease risk and key ovarian function markers—supporting clinical and public health action.