Approximately 10–20% of couples in the world are infertile (1). In China, the estimated infertility prevalence reaches 15.5% in the reproductive-aged female population (2). The most common causes of infertility are ovulatory dysfunction, male factor infertility, and tubal disease (3). Diminished ovarian reserve (DOR), a reflection of reproductive potential, is a significant cause of female infertility, accounting for about 10% of infertility cases (4). Ovarian reserve refers to the quantity and quality of the remaining oocytes in the ovaries. DOR describes women of reproductive age having menses whose response to ovarian stimulation or fecundity is reduced compared with women of comparable age (3, 5).

In recent years, DOR prevalence has increased and now tends to occur at younger ages, with the reported prevalence of DOR varying between 10 and 35%, depending on differences in the definition of DOR (7). DOR not only leads to reproductive dysfunctions such as menstrual irregularities, recurrent miscarriage, and infertility, but may also result in premature menopause, thereby adversely affecting women’s quality of life (8, 9).

Although age has been established as a well-known independent risk factor, the etiology of DOR remains elusive (6, 10). As a complex clinical condition, ovarian reserve can be influenced by a range of factors, including medical causes, genetic variables, environmental exposures, and unidentified contributors (6, 11). With an increasing number of women delaying marriage and childbirth, DOR has emerged as an intractable and complex issue for women wishing to conceive, as well as a broader societal challenge for future generations (12). While previous studies have primarily focused on ovarian reserve markers such as anti-Müllerian hormone (AMH) and antral follicle count (AFC) (13–15), limited research has used DOR as a primary outcome, and the potential association between TORCH infections and DOR risk remains unclear.

To address this knowledge gap, we conducted a large-scale, age-matched case–control study to comprehensively characterize the clinical features of DOR and identify potential risk factors, including TORCH infections (toxoplasmosis, others, rubella, cytomegalovirus, herpes) and DOR risk remains unclear.

To our knowledge, this is the first large-scale, age-matched case-control study investigating DOR in Chinese women. We identified significant associations between DOR and factors including ethnicity, occupation, obesity, light menstrual flow, two or more pregnancies, and T. gondii infection. Notably, the association between pregnancies and infection with T. gondii, CMV, and RV infections with DOR varied across age groups, with evidence of significant effect modification by age. These findings suggest new links between TORCH infections and DOR, and highlight the need for age-specific prevention strategies.

FSH, AFC, and AMH are commonly used in ovarian reserve examinations. AMH is primarily secreted by primary, preantral, and early antral follicles. With advancing age, the number of ovarian follicles declines, leading to decreased concentrations of AMH (5, 21). FSH secretion is regulated by the hypothalamic–pituitary–ovarian (HPO) axis. Under normal ovarian function, sufficient levels of E2 and inhibin B are produced to exert negative feedback on FSH secretion, maintaining it within the normal range. Elevated FSH levels indicate DOR, suggesting impaired ovarian feedback and insufficient hormonal suppression of FSH (5). Importantly, measurement of both FSH and E2 on cycle day 3 may help decrease the incidence of false-negative testing (6). AFC is the sum of the number of antral follicles in both ovaries during the early-follicular phase. with a higher count generally indicating preserved ovarian function. In our study, significant differences were determined in all measures of ovarian reserve between two groups, and DOR group exhibited significantly lower AMH, AFC, and T levels, and higher FSH, LH and E2 levels. In women with DOR, impaired negative feedback regulation may lead to elevated LH secretion. The reduction in follicle number and the decline in the functional capacity of the theca cells may result in decreased testosterone levels. There were no significant differences in progesterone levels measured during the early follicular phase. This finding is biologically plausible, as progesterone secretion during this phase is minimal and primarily reflects the absence of luteal activity rather than ovarian reserve status. Unlike anti-Müllerian hormone (AMH) and follicle-stimulating hormone (FSH), which directly reflect follicular quantity and responsiveness, basal progesterone is not considered a sensitive marker of diminished ovarian reserve (6, 22). In addition, consistent with previous studies (23), participants with DOR exhibited lower serum PRL levels, which may reflect an overall alteration in the function of the HPO axis. This suggests that lower prolactin levels are likely a feedback result of diminished ovarian reserve or a reduced follicular pool, rather than a causative factor. PRL secretion is influenced by various factors such as emotions and stress (24). The present study excluded patients with hyperprolactinemia, and the observed differences fell within the low end of the normal reference range. This indicates that when evaluating ovarian reserve, attention should not only be paid to elevated prolactin levels; lower levels may also be associated with ovarian functional status.

Data from Carlos I et al. indicate that ethnicity may influence ovarian reserve, with Latina and Chinese women exhibiting lower AMH levels compared to White women, suggesting DOR and a higher risk of earlier menopause (25, 26). Our findings extend this observation, demonstrating that ethnic disparities may exist even within the same racial group (East Asians), as evidenced by differences between Han and minority populations in China. Further longitudinal studies are needed to clarify whether these differences are driven by genetic, nutritional, environmental, or lifestyle factors, and to elucidate the underlying mechanisms for better prediction of reproductive potential and long-term health outcomes. Ethnicity also modulates the association between obesity and ovarian reserve. Compared to Asian and Hispanic women, White women exhibit higher baseline AMH levels, with elevated BMI further reducing AMH levels in White and African-American women but not in Asian or Hispanic populations (13, 27, 28). Notably, variations in obesity classification criteria may contribute to these disparities. While BMI is widely used to assess health risks, its optimal thresholds differ across ethnic groups. For a given BMI, non-Hispanic Asians have more body fat than non-Hispanic Whites (29). In our study, adopting the Chinese BMI criteria (obese defined as BMI ≥28 kg/m²), we identified obesity as a significant risk factor for DOR, consistent with findings reported by Li YL et al. and Moslehi N et al. (30, 31). Regular physical activity helps to mitigate the tendency for weight gain and adverse changes in body composition and fat distribution that accompany aging and the menopausal transition (32). However, women engaging in high-intensity physical activity or demanding occupational labor exhibit reduced fertility, manifested by fewer retrieved mature oocytes following controlled ovarian hyperstimulation, as demonstrated in our study (33).

Previous studies have identified light menstrual flow as a risk factor for infertility, whereas number of pregnancies serves as a protective factor (2, 34). Our findings demonstrate similar associations between menstrual flow, pregnancy, and DOR. In addition, a study investigating the effects of reproductive and lifestyle factors on age-specific AMH levels reported that pregnant women exhibited significantly lower AMH concentrations, while higher parity was associated with higher AMH (35). Currently, no conclusive evidence suggests that pregnancy accelerates or slows the long-term decline of ovarian reserve. Whether multiple pregnancies exert a protective effect on ovarian reserve by reducing the number of ovulatory cycles and subsequent follicle depletion, or whether the observed association reflects a reverse causality—where DOR leads to reduced fertility and thus fewer pregnancies—remains uncertain and warrants cautious interpretation. Light menstrual flow may also represent a clinical manifestation of advanced ovarian reserve decline. DOR can lead to a shortened follicular phase, reduced estrogen peak levels, inadequate endometrial proliferation, and consequently, decreased menstrual volume. However, due to study design limitations, causal inferences remain constrained, warranting further prospective research to elucidate the underlying mechanisms. Nevertheless, persistent reductions in menstrual flow in women under 35 years should raise clinical suspicion for premature ovarian insufficiency or early ovarian aging, necessitating prompt ovarian reserve assessment.

TORCH infections are known to cause adverse pregnancy outcomes such as miscarriage, preterm birth, and congenital anomalies (36, 37). In our study, the seroprevalence of T. gondii (11.4%) was higher than that reported in the general population of women of reproductive age. Given that the participants were from Southwest China—a region with the highest ethnic diversity in the country—this finding may reflect a distinct regional epidemiological context, including local dietary practices, environmental exposures, and pathogen circulation. In addition, this elevated prevalence may partly relate to the characteristics of the study population, which consisted of women seeking fertility evaluation and therefore may differ from the general population. Our findings reveal a significant association between T. gondii infection and DOR. Notably, stratified analyses by age exhibited TORCH infections (including T. gondii, CMV and RV) significantly related to DOR among aged <35 years subgroup, whereas no significant correlation was observed in older women. This age-dependent interaction was significant, possibly because the physiological decline in ovarian reserve with age overshadows the effects of infection. Additionally, younger women may exhibit a more pronounced or sustained immune response to infections, leading to greater ovarian tissue damage. However, it remains unclear whether the observed associations are attributable to primary infections or to reactivation. Notably, the very large odds ratios observed for TORCH infections, particularly T. gondii (OR = 23.750), in women aged 20–35 years were accompanied by wide confidence intervals. This likely reflects sparse-data bias arising from the small number of exposed controls in this subgroup, resulting in limited precision and instability of the effect estimates. Accordingly, these findings should be interpreted as signals of potential association rather than as precise estimates of effect magnitude. In BMI-stratified analysis, the association between TORCH infection and DOR was evident among BMI ≤24 kg/m² subgroup; however, no significant interaction was detected, indicating that BMI may not substantially modify this relationship. TORCH infections may contribute to DOR through immune activation, inflammatory responses, and direct ovarian tissue damage (38, 39). Whilst HSV is renowned for its tropism towards genital tract epithelium and can cause pelvic inflammatory disease, its primary pathology is typically focal and recurrent (40). In contrast, pathogens such as Toxoplasma gondii or rubella virus cause systemic infections, potentially exerting broader immunopathological or direct cytotoxic effects on ovarian tissue. Alternatively, the lack of association may stem from timing issues in our assessment; common HSV reactivation may not be adequately captured by a single IgM test, or its impact on ovarian reserve may be subtle. Overall, TORCH screening may be considered for younger women with unexplained DOR, particularly in regions with a high prevalence of these infections. However, it must be emphasized that our results do not provide evidence that antimicrobial or antiviral treatment targeting these infections can reverse or improve ovarian reserve. Whether pathogen-specific interventions can modify ovarian function remains a fundamental question that warrants mechanistic investigations and prospective epidemiological studies.

The key strengths of this study include the adequate sample size and age-matched case-control design, which enhanced study efficiency. Nonetheless, several limitations should be acknowledged. First, residual confounding may persist due to the lack of data on environmental exposures, dietary habits, and nutritional supplementation. Second, the assessment of certain clinical variables, such as “menstrual flow”, was based on patient self-report documented in electronic medical records without a standardized quantitative definition. This reliance on subjective recall introduces the potential for misclassification and recall bias, which may have attenuated the observed associations for this factor. Third, TORCH infections in this study were assessed based on serological IgM positivity. Although IgM testing increases specificity for recent or active infection, it does not allow precise determination of the timing of exposure and may also reflect persistent antibodies or pathogen reactivation. In the absence of longitudinal serological follow-up or documented infection history, the biological interpretation of infection-related ovarian impairment should therefore be considered speculative. Additionally, since all participants were recruited from a maternity hospital, and the control group included women with underlying gynecological conditions such as tubal obstruction and endometrial polyps, caution should be exercised when generalizing the findings to the general population. Due to the observational nature of the study, additional limitations such as multiple testing, measurement bias, and temporal ambiguity cannot be ruled out. Future large-scale prospective cohort or experimental studies are needed to further investigate the determinants and underlying mechanisms of ovarian reserve and to validate our findings.

In summary, this large age-matched case-control study identified multiple clinical and demographic factors associated with DOR, including ethnicity, obesity, menstrual flow, pregnancy history, and TORCH infections, with evidence of age-related effect modification for pregnancy history and infection. While the study provides novel insights into risk stratification, limitations such as potential residual confounding and diagnostic variability highlight the need for standardized DOR criteria and prospective investigations. Future research should prioritize elucidating causal mechanisms—particularly the role of infections and immune responses—and validate these associations in diverse populations to inform clinical screening and preventive strategies for at-risk women.