Moderating effect of dysmenorrhea on the association between number of pregnancies and antenatal depression in the third trimester: a cross-sectional study.

A total of 1,178 pregnant women aged 18 to 46 years were included in the study (Table  1 ), with an average age of 29.49 years (SD = 3.96). Preliminary analysis indicated no significant association between gestational week and EPDS scores ( p  > 0.05). Descriptive statistics for the entire sample, including variables such as age, pre-pregnancy BMI, number of pregnancies, number of deliveries, and EPDS scores, are presented first. These data provide an overview of the sample’s demographic and clinical characteristics. The general clinical data comparison between the two groups is presented in Fig.  1 . Table 1 Characteristics of the full sample and by dysmenorrhea status Variables All ( n  = 1178) Non-dysmenorrhea ( n  = 624) Dysmenorrhea ( n  = 554) F/χ² p Age (years) 29.49 ± 3.96 30.10 ± 3.98 28.80 ± 3.82 32.5 < 0.001 Education (%) 1.02 0.796  Primary school /Junior school 102 (8.7) 51 (4.3) 51 (4.3)  Technical secondary school /High school 188 (16.0) 97 (8.2) 91 (7.7)  Junior college /Undergraduate 827 (70.2) 441 (37.4) 386 (32.8)  Graduate school and above 61 (5.2) 35 (3.0) 26 (2.2) Pre-gestation BMI (kg/m²) 20.86 ± 3.47 21.11 ± 3.70 20.58 ± 3.17 6.89 0.009 Marital status (%) 8.39 0.004  Unmarried 67 (5.7) 24 (2.0) 43 (3.7)  Married 1111 (94.3) 600 (50.9) 511 (43.4) Number of pregnancies (times) 1.91 ± 1.11 2.03 ± 1.18 1.77 ± 1.01 -3.93 < 0.001 Number of deliveries (times) 0.35 ± 0.53 0.44 ± 0.58 0.25 ± 0.46 -5.74 < 0.001 Husband Education (%) 0.91 0.822  Primary school /Junior school 96 (8.1) 52 (4.4) 44 (3.7)  Technical secondary school /High school 191 (16.2) 106 (9.0) 85 (7.2)  Junior college /Undergraduate 803 (68.2) 422 (35.8) 381 (32.3)  Graduate school and above 88 (7.5) 44 (3.7) 44 (3.7) EPDS scores 7.80 ± 5.77 6.97 ± 5.32 8.77 ± 6.11 -5.24 < 0.001 Variables using percentage are reported as a chi-square test between with dysmenorrhea and without dysmenorrhea groups. Number of pregnancies, deliveries and EPDS scores were compared using the Man-Whitney rank sum test, and other continuous variables were compared using a one-way ANOVA test. Means ± standard deviation (SD) was employed to describe continuous variables, and frequency and percentage were utilized to depict categorical variables Abbreviations : BMI Body mass index, EPDS Edinburgh postnatal depression scale Characteristics of the full sample and by dysmenorrhea status Variables using percentage are reported as a chi-square test between with dysmenorrhea and without dysmenorrhea groups. Number of pregnancies, deliveries and EPDS scores were compared using the Man-Whitney rank sum test, and other continuous variables were compared using a one-way ANOVA test. Means ± standard deviation (SD) was employed to describe continuous variables, and frequency and percentage were utilized to depict categorical variables Abbreviations : BMI Body mass index, EPDS Edinburgh postnatal depression scale Fig. 1 Shows the comparison of various clinical variables between the non-dysmenorrhea and dysmenorrhea groups. Panel A shows the comparison of age, Panel B shows pre-gestation BMI, Panel C presents the number of pregnancies, Panel D presents the number of deliveries, and Panel E shows EPDS scores. Significant differences between groups are indicated with asterisks (*** p  < 0.001, ** p  < 0.01) Shows the comparison of various clinical variables between the non-dysmenorrhea and dysmenorrhea groups. Panel A shows the comparison of age, Panel B shows pre-gestation BMI, Panel C presents the number of pregnancies, Panel D presents the number of deliveries, and Panel E shows EPDS scores. Significant differences between groups are indicated with asterisks (*** p  < 0.001, ** p  < 0.01) Following this, stratified analyses were performed based on dysmenorrhea status to assess the moderating effect of dysmenorrhea on the relationship between the number of pregnancies and EPDS scores, as described in the research hypothesis. The non-dysmenorrhea group exhibited a higher mean age (30.10 ± 3.98 years) compared to the dysmenorrhea group (28.80 ± 3.82 years), with a statistically significant difference ( p  < 0.001, Fig.  1 A). The pre-pregnancy BMI was higher in the non-dysmenorrhea group compared to the dysmenorrhea group ( p  < 0.01, Fig.  1 B). Additionally, the number of pregnancies and deliveries was greater in the non-dysmenorrhea group ( p  < 0.001, Fig.  1 C and D). The EPDS score was significantly lower in the non-dysmenorrhea group than in the dysmenorrhea group ( p  < 0.001, Fig.  1 E). Marital status analysis showed that a significantly higher proportion of unmarried individuals were found in the dysmenorrhea group ( p  = 0.004). Pearson correlation analysis revealed that the education levels of both the pregnant women and their husbands were negatively associated with EPDS scores in both groups (non-dysmenorrhea group: r =-0.14, p  < 0.001 and r =-0.16, p  < 0.001; dysmenorrhea group: r =-0.12, p  < 0.01 and r =-0.15, p  < 0.001, Table  2 ). In the non-dysmenorrhea group, a significant negative correlation existed between age and EPDS scores ( r =-0.19, p  < 0.001), as illustrated in Fig.  2 A. Within the dysmenorrhea group, a positive correlation was identified between the number of pregnancies and EPDS scores ( r  = 0.14, p  < 0.01), as well as between the number of deliveries and EPDS scores ( r  = 0.08, p  < 0.05), as depicted in Fig.  2 B. Table 2 Correlation analysis between baseline characteristics Items Age Education Progestation BMI Marital status Number of pregnancies Number of deliveries Husband Education EPDS scores Age - 0.32*** 0.18*** 0.19*** 0.34*** 0.40*** 0.23*** -0.07 Education 0.15*** - 0.11* 0.11* -0.17*** -0.13** 0.52*** -0.12** Progestation BMI 0.11** -0.12** - 0.08 0.06 0.12** 0.02 0.03 Marital status 0.16*** 0.02 0.08 - 0.03 0.10* 0.16*** -0.01 Number of pregnancies 0.37*** -0.24*** 0.09* 0.06 - 0.64*** -0.13** 0.14** Number of deliveries 0.41*** -0.20*** 0.11** 0.14*** 0.63*** - -0.07 0.08* Husband Education 0.19*** 0.56*** -0.10** 0.07 -0.19*** -0.16*** - -0.15*** EPDS scores -0.19*** -0.14*** 0.02 -0.05 0.04 -0.01 -0.16*** - All data were reported as correlation analysis Abbreviations : EPDS , Edinburgh Postnatal Depression Scale, BMI Body Mass Index * p  < 0.05, ** p  < 0.01, *** p  < 0.001. Values above the diagonal represent correlation coefficients for the dysmenorrhea group; values below the diagonal represent correlation coefficients for the non-dysmenorrhea group Correlation analysis between baseline characteristics All data were reported as correlation analysis Abbreviations : EPDS , Edinburgh Postnatal Depression Scale, BMI Body Mass Index * p  < 0.05, ** p  < 0.01, *** p  < 0.001. Values above the diagonal represent correlation coefficients for the dysmenorrhea group; values below the diagonal represent correlation coefficients for the non-dysmenorrhea group Fig. 2 Correlation analysis of EPDS scores. Correlation between EPDS scores and factors such as age, number of pregnancies, number of deliveries, marital status, pre-pregnancy body mass index (BMI), education level of the pregnant woman and her husband in dysmenorrhea ( b ) and non-dysmenorrhea ( a ) groups Correlation analysis of EPDS scores. Correlation between EPDS scores and factors such as age, number of pregnancies, number of deliveries, marital status, pre-pregnancy body mass index (BMI), education level of the pregnant woman and her husband in dysmenorrhea ( b ) and non-dysmenorrhea ( a ) groups Before running the regression models, we assessed multicollinearity among the predictors by calculating the VIF. The correlation between the number of pregnancies and the number of deliveries was strong ( r  = 0.63–0.64), indicating the possibility of collinearity. The VIF values for these two variables were calculated: Number of pregnancies: VIF = 4.2; Number of deliveries: VIF = 4.1. Since both VIF values were below the threshold of 5, both variables were retained in the regression models. Next, we conducted linear regression analyses to assess the moderating effect of dysmenorrhea on EPDS scores. The results are summarized in Table  3 . Age (β = -0.15, t = -4.45, p  < 0.001) and the education level of the husbands (β = -0.09, t = -2.56, p  < 0.05) were negatively associated with EPDS scores. Dysmenorrhea was positively associated with EPDS scores (β = 0.15, t = 5.24, p  < 0.001). Moreover, the interaction effect between the number of pregnancies and the presence of dysmenorrhea also showed a positive association with EPDS scores (β = 0.08, t = 2.22, p  < 0.05). Table 3 Linear regression for the moderating effect of dysmenorrhea on EPDS scores Variables Model 1 (β) Model 2 (β) Model 3 (β) VIF ΔR² (from Model 2 to Model 3) Age -0.15*** -0.15*** -0.15*** 1.5 Education -0.05 -0.03 -0.03 2 Progestation BMI 0.02 0.03 0.03 1.8 Marital status 0.00 0.01 0.01 1.3 Number of deliveries 0.04 0.00 0.00 4.1 Number of pregnancies - 0.12** 0.07 4.2 0.01 Husband Education -0.09* -0.09* -0.09* 2.5 Without/With dysmenorrhea - 0.15*** 0.15*** 1.6 Number of pregnancies×Without/With dysmenorrhea - - 0.08* 4 The VIF values are based on the actual VIF calculations from your regression models. If any variable has a VIF greater than 5, consider excluding it or combining it with another variable (e.g., create a composite variable) ΔR² from Model 2 to Model 3 = 0.01, representing a small effect Abbreviations: EPDS Edinburgh Postnatal Depression Scale, BMI Body Mass Index * p  < 0.05, ** p  < 0.01, *** p  < 0.001 Linear regression for the moderating effect of dysmenorrhea on EPDS scores The VIF values are based on the actual VIF calculations from your regression models. If any variable has a VIF greater than 5, consider excluding it or combining it with another variable (e.g., create a composite variable) ΔR² from Model 2 to Model 3 = 0.01, representing a small effect Abbreviations: EPDS Edinburgh Postnatal Depression Scale, BMI Body Mass Index * p  < 0.05, ** p  < 0.01, *** p  < 0.001 The moderation analysis reports beta coefficients, t-values, and p-values, along with standardized effect sizes. For example, the R² change from Model 2 to Model 3 (0.07 to 0.08, representing ΔR² = 0.01) suggests a small effect size, which should be interpreted in the context of this study. The addition of the interaction term between the number of pregnancies and dysmenorrhea has a small but significant contribution to explaining the variance in EPDS scores. To further explore this moderating effect, a simple slope analysis was performed, and an effect plot was generated. Figure  3 illustrates the moderating effect of dysmenorrhea on the relationship between the number of pregnancies and EPDS scores. Figure  4 depicts the effect of the number of pregnancies on EPDS scores in the non-dysmenorrhea and dysmenorrhea groups. In the non-dysmenorrhea group, EPDS scores gradually decreased as the number of pregnancies increased. In contrast, in the dysmenorrhea group, EPDS scores increased with an increasing number of pregnancies. Fig. 3 Moderating effect of dysmenorrhea on EPDS scores. Linear regression analysis demonstrating the interaction effect of dysmenorrhea on the relationship between number of pregnancies and EPDS scores Moderating effect of dysmenorrhea on EPDS scores. Linear regression analysis demonstrating the interaction effect of dysmenorrhea on the relationship between number of pregnancies and EPDS scores Fig. 4 Effect plot of dysmenorrhea moderation. Visualization of the moderating role of dysmenorrhea in the relationship between the number of pregnancies and EPDS scores Effect plot of dysmenorrhea moderation. Visualization of the moderating role of dysmenorrhea in the relationship between the number of pregnancies and EPDS scores

To test the hypothesis that dysmenorrhea moderates the association between the number of pregnancies and AND, we employed a moderated regression analysis framework within a cross-sectional observational design. This was a cross-sectional observational study involving pregnant women in their third trimester. Sample size was determined a priori using G*Power 3.1 for a linear multiple regression with a small-to-medium interaction effect (f² = 0.02), α = 0.05, power = 0.80, and 8 predictors (including the interaction term). This yielded a required sample size of at least 395 participants. To account for potential dropouts and ensure robust subgroup analyses, we planned to enroll at least 1,200 participants. A total of 1,191 women in their third trimester of pregnancy who attended the Second Affiliated Hospital of Xiamen Medical College were enrolled from March 2017 to May 2019. After excluding missing samples, 1,178 pregnant women were considered in the final analytical dataset. Inclusion criteria: (1) Women in the third trimester of pregnancy (defined as gestational age 28–41 weeks) at the time of questionnaire administration, regardless of whether delivery subsequently occurred preterm. The EPDS scores were examined for systematic variation across the third trimester, with specific attention to any differences by gestational week; (2) pregnant women who had singleton pregnancy; (3) aged ≥ 18 years; (4) no fetal abnormalities confirmed by prenatal examination; (5) able to read and correctly understand the questionnaire. Exclusion criteria: (1) drug or alcohol abusers; (2) pregnant women with a documented history of disorders, including mental retardation or psychiatric conditions; (3) infectious diseases; (4) severe internal and surgical diseases or serious complications during pregnancy; (5) other serious organic lesions. Based on responses to a structured questionnaire, the presence of dysmenorrhea was determined by asking participants whether they experienced menstrual pain during their cycles prior to pregnancy(the 12 months preceding pregnancy). Participants who reported experiencing menstrual pain were categorized into the dysmenorrhea group, while those who reported no pain constituted the non-dysmenorrhea group. Participants were then categorized into two groups: those without dysmenorrhea and those with dysmenorrhea. Clinical and demographic information was collected from all participants using a structured questionnaire administered in person during routine prenatal visits at the obstetrics outpatient clinic of the Second Affiliated Hospital of Xiamen Medical College. Data collection was conducted by trained obstetric nurses or research assistants who had received standardized instruction on questionnaire administration. The survey included items on maternal age, marital status, number of pregnancies and deliveries, pre-pregnancy body mass index (BMI), education level of the pregnant woman and her husband, and the presence or absence of dysmenorrhea (based on self-report). The questionnaire was finished by the participants in a private setting in the hospital, with research staff available to clarify any questions if needed. All questionnaires were completed between 28 and 41 weeks of gestation and before delivery. The number of pregnancies was assessed based on gravidity, representing the total number of prior pregnancies regardless of outcome. Data on multiple gestations (e.g., twin or triplet pregnancies) were not collected or analyzed as part of this study. The EPDS was utilized to evaluate the occurrence of depressive symptoms among pregnant women in the month leading up to delivery [ 11 , 12 ]. The scale consisted of 10 items, scoring each question on a scale of 0 to 3, reflecting the degree of symptom severity. Higher scores indicated greater levels of depression. The EPDS has demonstrated good internal consistency in previous studies, with Cronbach’s alpha ranging from 0.76 to 0.93 [ 13 ]. Consistent with validated Chinese criteria, an EPDS score exceeding 10 was utilized to indicate probable AND warranting clinical evaluation [ 14 ]. In a validation study of 534 Chinese pregnant women, this cutoff demonstrated a sensitivity of 0.82 (95% CI: 0.72–0.89), specificity of 0.86 (95% CI: 0.83–0.89), and a positive predictive value of 0.43 [ 14 ]. It is important to acknowledge that the EPDS functions as a screening instrument rather than a diagnostic tool; scores indicate probable depression that requires clinical confirmation. In our study, the internal consistency of the EPDS was good, with Cronbach’s alpha of 0.89. Linear regression analysis was used to assess the moderating role of dysmenorrhea. First, all of the continuous variables were normalized. EPDS score was used as the dependent variable. Maternal age, education level, pre-gestation BMI, marital status [ 13 , 15 ], number of deliveries, and husband’s education level were used as covariates in Model 1 [ 16 , 17 ], based on theoretical considerations and empirical evidence suggesting their relevance to AND. Model 2 incorporated the independent variable of the number of pregnancies and the moderating variable of dysmenorrhea status based on Model 1. Model 3 added the interaction term between the independent variable of the number of pregnancies and the moderating variable of dysmenorrhea status based on Model 2. Hierarchical multiple regression analysis was conducted with covariates to analyze the moderating effects after regulating for education level, number of pregnancies, age, pre-pregnancy BMI, marital status, and husband’s education level. Simple slope analysis was further employed to evaluate these moderating effects. Data was analyzed using SPSS 22.0, R (4.3.1) software, and GraphPad Prism 7.0. The mean ± SD was reported for continuous variables, and categorical variables were depicted in frequency and percentage. Variance homogeneity was assessed for age and pre-pregnancy BMI, with one-way ANOVA applied to discern differences across the groups. The Mann-Whitney U test was utilized to compare the remaining continuous variables, and the chi-square test was applied to analyze categorical variables. Subsequently, Pearson correlation analysis was performed to examine the relationship between EPDS scores and continuous demographic variables, such as maternal age and number of pregnancies. Categorical variables were analyzed separately using appropriate non-parametric or comparative tests. All tests were two-tailed, with statistical significance set at p  < 0.05. The research received clearance from the Ethics Committee at Xiamen Medical College’s Second Affiliated Hospital. In line with the ethical standards of the Declaration of Helsinki, the study secured written informed consent from each participant or, where applicable, their legal guardians.

Our study identified a significant positive association between the number of pregnancies and AND, with dysmenorrhea further exacerbating this relationship. These findings offer valuable theoretical insights for the fields of psychiatry and perinatal medicine. Future research should aim to clarify the underlying mechanisms linking reproductive history and menstrual pain with antenatal mental health, ideally through prospective cohort studies that incorporate hormonal, inflammatory, and psychosocial factors. Additionally, investigating the potential mediating role of chronic gynecological conditions such as endometriosis could help refine risk stratification and intervention strategies.

In recent years, AND has emerged as a prevalent psychological disorder during pregnancy. The primary symptoms include anxiety and extreme irritability, and the mental state of pregnant women directly influences the delivery process and fetal conditions, often resulting in poor outcomes such as prolonged labor and neonatal asphyxia [ 4 ]. Pregnant women with severe AND frequently experience hyperemesis gravidarum, which could lead to preterm birth and miscarriage [ 18 , 19 ]. Therefore, early screening of high-risk pregnant women is deemed essential for maternal and child healthcare. However, research on the early screening of AND remains limited, lacking adequate theoretical support. Our study discovered a noteworthy positive association between the number of pregnancies and AND, with dysmenorrhea further exacerbating this correlation. This finding indicates that during early prenatal examinations, especially for those with a history of higher number of pregnancies, a history of dysmenorrhea might increase the risk of AND. This study innovatively highlights the moderating influence of dysmenorrhea in the relationship between the number of pregnancies and AND. Menstrual pain, known as dysmenorrhea, is estimated to affect 16% to 91% of women, with severe pain experienced by 2% to 29% of these cases [ 20 ]. Our findings indicated that the non-dysmenorrhea group comprised younger individuals than the dysmenorrhea group. Charumathi et al. conducted a cross-sectional analytical study within a community framework containing 246 women in the reproductive age range. They also found that age was significantly positively correlated with dysmenorrhea, and young women were more likely to experience dysmenorrhea [ 21 ]. Additionally, our study revealed that the pre-pregnancy BMI was elevated in the non-dysmenorrhea group relative to the dysmenorrhea group. In a parallel vein, a cross-sectional investigation into the link between dietary practices and menstrual irregularities in women indicated that women without menstrual disorders have a higher BMI compared to those with such issues [ 22 ]. Moreover, our results indicated that the number of pregnancies and deliveries was higher in the non-dysmenorrhea group compared with the dysmenorrhea group. While we did not assess endometriosis (EM) status in this study, it is noteworthy that EM is a chronic gynecologic condition in which dysmenorrhea is a predominant clinical symptom [ 23 ]. Previous literature suggests that fertility may be impaired in women with EM, contributing to reduced pregnancy and delivery rates [ 24 ]. Therefore, it is possible that underlying reproductive conditions such as EM may partially explain this observed difference, although further studies are needed to confirm this association in our population. Our findings revealed that the education levels of both the pregnant women and their husbands were negatively correlated with EPDS scores in neither non-dysmenorrhea nor dysmenorrhea women. The extant literature presents equivocal findings regarding the education-AND relationship, potentially reflecting contextual factors such as socio-economic status, cultural differences, or coping mechanisms [ 25 , 26 ]. The present findings suggest that higher education levels may be associated with a lower risk of AND, which could be attributed to better access to resources and coping strategies, though further investigation is needed to clarify this relationship. The present study suggests that the relationship between education and AND may be more complex, influenced by factors such as marital status, socio-economic pressures, and individual coping strategies. These findings indicate the need for further investigation into how education level interacts with other psychosocial factors to influence prenatal mental health, and they provide a basis for developing targeted interventions. As higher education proliferates in China and women’s employment prospects grow, individuals with higher education levels are increasingly educated women, and men may face more job opportunities and challenges. They support a part of family income, so they should not only adapt to psychosomatic changes during pregnancy but also bear the pressure of family and work, causing the increased risk of AND. In this study, age was negatively correlated with EPDS scores in non-dysmenorrhea pregnant women. The observed increased risk of AND in younger women may be attributed to a range of factors, including psychosocial vulnerability, less-established coping mechanisms, and potentially lower levels of social support. In contrast, older pregnant women often benefit from accumulated coping strategies, established support networks, and greater life experience, which may help mitigate the risk of depression during pregnancy [ 27 ]. However, the mechanisms underlying age-related differences in AND susceptibility warrant further investigation, particularly in relation to how social, psychological, and biological factors interact across the lifespan. Additionally, for the dysmenorrhea pregnant women, the number of pregnancies and the number of deliveries were connected with EPDS scores. The possible reason is that some adverse events during pregnancy have a psychological shadow on women, resulting in corresponding anxiety and depression during re-pregnancy. It may also be related to possible dystocia or abortion. Our research revealed a significant positive correlation between the number of pregnancies and the incidence of prenatal depression, and dysmenorrhea further exacerbated this relationship. This result is consistent with the existing literature. First, pregnant women with a higher number of pregnancies may experience increased psychological and physical burdens. Each pregnancy brings physical changes and psychological stress [ 28 ]. Secondly, from a mechanistic perspective, an increasing number of pregnancies may contribute to a heightened inflammatory state. Normal higher gravidity are associated with low-grade systemic inflammation, commonly indicated by elevated high-sensitivity C-reactive protein [ 29 ]. In contrast, pathological reproductive histories such as recurrent pregnancy loss are characterized by immune dysregulation, with increased levels of pro-inflammatory cytokines including IL-6 and TNF-alpha [ 30 , 31 ]. Accumulating evidence indicates that prenatal depressive symptoms are associated with elevated pro-inflammatory cytokines [ 32 , 33 ]. The inflammatory hypothesis, while biologically plausible, should be framed explicitly as hypothesis-generating speculation requiring empirical verification in the study population. Moreover, cytokines and other pro-inflammatory factors mediate primary dysmenorrhea [ 34 ]. The increase in these inflammatory factors is closely related to the occurrence of depression in pregnant women [ 35 , 36 ]. Although not directly examined in the current investigation, dysmenorrhea may plausibly increase the risk of AND through several putative molecular mechanisms that warrant empirical evaluation in future research. Dysmenorrhea, as a chronic pain symptom, can trigger a long-term inflammatory response, which is closely related to the pathogenesis of depression [ 34 ]. Menstrual pain frequently co-occurs with elevated inflammatory factors, including prostaglandin E2 and IL-6, which are elevated in dysmenorrhea and depression [ 34 ]. Inflammatory factors, such as 5-HT (5-hydroxytryptamine, serotonin) and dopamine, affect the neurotransmitter system by activating the inflammatory pathway in the body, thus affecting emotional regulation and psychological state [ 37 ]. In addition, dysmenorrhea may increase the risk of prenatal depression by affecting the neuroendocrine system. Persistent pain and stress can trigger the hypothalamic-pituitary-adrenal (HPA) axis, increasing cortisol levels [ 38 ]. Elevated cortisol levels can impact the brain regions responsible for emotional regulation, such as the hippocampus and prefrontal cortex, thereby increasing the incidence of depressive symptoms [ 38 ]. Furthermore, emerging studies suggest that the mode of delivery and breastfeeding practices may also influence maternal mental health [ 39 , 40 ]. Cesarean delivery, especially when unplanned, has been associated with a higher risk of postpartum depression and may reflect an underlying vulnerability to antenatal distress. Similarly, breastfeeding has been shown to exert protective effects on mood via oxytocin release and enhanced maternal-infant bonding [ 41 ]. However, women with higher number of pregnancies may experience reduced breastfeeding success or increased cesarean rates, which could compound psychological stress and contribute to depressive symptoms. Future studies should examine these perinatal factors in conjunction with number of pregnancies and dysmenorrhea to elucidate their combined effects on antenatal mental health. This study has several notable strengths. First, it includes a relatively large sample size of pregnant women in the third trimester, enhancing the statistical power and generalizability of the findings within the regional population. Second, it explores the novel interaction between number of pregnancies and dysmenorrhea in predicting AND, an area that has received limited attention in existing literature. Third, the use of a validated screening tool—the EPDS—adds to the methodological robustness of the study. However, several limitations must be acknowledged to appropriately contextualize the findings. (1) Study Design and Sampling. First, the use of single-center convenience sampling may introduce selection bias and limit the external validity of the results to broader populations, particularly those with different healthcare-seeking patterns or socioeconomic profiles. Second, the cross-sectional design precludes causal or temporal inference regarding the relationships among the number of pregnancies, dysmenorrhea, and antenatal depression. Although this design was a necessary first step to identify a novel interaction effect, longitudinal studies are required to establish directionality and clarify potential causal pathways. (2) Measurement Limitations of Key Variables. Dysmenorrhea was assessed retrospectively via a single binary self-report item, which is vulnerable to recall bias and lacks information on pain severity, frequency, or etiology (primary vs. secondary). This precluded dose-response analyses and may have led to misclassification. Pre-pregnancy depression history was ascertained through a simple self-reported question rather than a validated diagnostic interview, likely underestimating its prevalence and limiting our ability to control for this established confounder. Moreover, we did not screen for endometriosis or other gynecological conditions that could independently influence both dysmenorrhea and depressive symptoms, potentially confounding the observed association. (3) Unmeasured Confounders and Psychosocial Factors. Important psychosocial determinants of antenatal mental health—such as partner support, relationship quality, and recent stressful life events—were not assessed, although they are known to moderate depression risk. Our definition of marital status was restricted to legal registration and did not capture cohabiting or stable unmarried partnerships, which may have led to an underestimation of social support. Additionally, several clinically relevant perinatal variables—including mode of delivery, breastfeeding intention, and postpartum outcomes—were not examined but could interact with pregnancy history and dysmenorrhea to shape maternal psychological well-being. (4) Consideration of Gestational Age. This study did not include gestational age as a covariate, its broad range and the potential for unmeasured physiological or psychological changes across the third trimester mean that residual confounding by gestational stage cannot be entirely ruled out. In summary, while this study provides novel evidence for the moderating role of dysmenorrhea in the relationship between gravidity and antenatal depression, the findings should be interpreted in light of these methodological constraints. Future research should employ multi-center prospective designs, incorporate validated, graded measures of dysmenorrhea and psychiatric history, conduct clinical screening for gynecological comorbidities, and include a broader array of psychosocial and perinatal covariates to better elucidate the underlying mechanisms and strengthen causal inference.

Depression during the perinatal period encompasses the manifestation of either severe or mild depressive episodes during the prenatal phase (antenatal depression) as well as within the first 12 months postpartum (postnatal depression) [ 1 ]. As the most common mental illness during pregnancy, the incidence of antenatal depression (AND) is reported to range from 15% to 65% worldwide [ 2 ]. A meta-analysis based on mainland China shows that the overall incidence of AND is about 20%, and women with a history of psychological or psychiatric conditions exhibit a higher prevalence [ 3 ]. It not only adversely affects maternal health but also impairs the cognitive development of infants and may even lead to suicidal tendencies in mothers or cause harm to infants, making it a major public health concern [ 4 ]. Presently, the known risk factors affecting the occurrence of AND include lower serum folic acid level, lower B12 level, personality traits, age, education level, social and family relationship status, and sleep of pregnant women [ 5 , 6 ]. Studies have indicated that the number of pregnancies and the presence of dysmenorrhea might play critical roles in the onset of AND. The National Health and Nutrition Examination Survey (NHANES) data reveal a significant positive association between the number of pregnancies and the risk of depression, with pregnant women demonstrating 1.52 times greater odds of depression (95% CI: 1.20–1.92) compared with the overall population after comprehensive covariate adjustment [ 7 ]. Nisar et al. identified a substantial correlation between prior pregnancies and AND [ 3 ]. A meta-analysis suggested a notable correlation between primary dysmenorrhea and depressive disorders [ 8 ]. Furthermore, Meng et al. observed an increased risk of AND associated with dysmenorrhea [ 9 ]. The Japanese JECS study also confirmed a significant association between dysmenorrhea and AND [ 10 ]. While previous research has indicated that the number of pregnancies and dysmenorrhea may independently impact AND, the interaction between these factors and their causal relationship with antenatal depression remains unclear. Therefore, drawing on the documented associations of gravidity and dysmenorrhea with inflammatory pathways and psychological stress, we hypothesize that dysmenorrhea moderates the positive link between the number of pregnancies and the risk of antenatal depression (AND). Specifically, we propose a conceptual model in which the relationship between higher gravidity and increased depressive symptoms is exacerbated among women with a history of dysmenorrhea, potentially due to cumulative inflammatory load or pain-related sensitization. This cross-sectional study was designed to test this moderating effect and establish a framework to elucidate the interplay between reproductive history and menstrual pain. Ultimately, our findings aim to provide insights for the development of targeted clinical screenings and effective interventions for AND.