This cross-sectional study was conducted at China’s Shanghai First Maternity and Infant Hospital. The participants were women who delivered in our hospital from January to April 2021 and completed the follow-up 42 days after delivery. We restricted the participants to women with singleton pregnancies who were 18–45 years old and had a good understanding of written and spoken Chinese. A total of 1,551 women were enrolled, including 147 mothers who had experienced lactational mastitis diagnosed by breast specialists during the postpartum period. Lactational mastitis is currently diagnosed by clinical evaluation based on signs of inflammation, pain, tenderness, heat, and engorgement, and the patients may experience flu-like symptoms (5, 12).

This study was approved by the Human Ethics Committee of the Shanghai First Maternity and Infant Hospital (KS20302) and conducted in accordance with the tenets of the Declaration of Helsinki. All participants signed informed consent.

Outcome measures included depression symptoms, which were evaluated using the Edinburgh Postnatal Depression Scale (EPDS) and the Patient Health Questionnaire 9 (PHQ9) at 6 weeks after delivery. The EPDS is a 10-item self-report questionnaire used explicitly for postpartum women (13). The Chinese version has concurrent validity, with a standardized Cronbach’s α of 0.73 (14). The presence of depression symptoms was defined as an EPDS score ≥ 9 (including minor and major depression), with a sensitivity of 85% and specificity of 82% (15). The PHQ9 is a 9-item self-report questionnaire that evaluates depression symptoms; the American College of Obstetricians and Gynecologists (ACOG) recommends administering PHQ9 to women during the perinatal period (4). The Chinese version has satisfactory concurrent validity, with a standardized Cronbach’s α of 0.86 (16). The presence of depression symptoms was defined as a PHQ9 score ≥ 5, with a sensitivity of 75% and specificity of 90% (4). The PHQ9 is widely used as a general screening tool for depression, including somatic symptoms such as sleep quality, which is considered a risk factor for the negative maternal mental disorder during the postpartum period (17), while the EPDS places less emphasis on somatic symptoms of depression. The degree of agreement between the EPDS and PHQ9 is only 0.5 during the postpartum period, with a discordance of 17% (18, 19). It has been suggested that both tools are categorized differently during the postpartum period (20). Therefore, the present study comprehensively evaluated the potential relationship between lactational mastitis and depression during the postpartum period, using both screening tools.

We selected covariates based on previous studies and expert knowledge (21, 22). A structured questionnaire was designed to collect the selected covariates, including maternal age (< 35 or ≥ 35 years), education level (junior high school or below, high school, bachelor’s degree or above), employment status during pregnancy (employed/unemployed), prenatal depression (yes/no), smoking during pregnancy (yes/no), alcohol use during pregnancy (yes/no), nulliparous (yes/no), maternal comorbidities (yes/no, including gestational diabetes mellitus, hypertension, and thyroid disease), delivery modality (cesarean or vaginal delivery), and neonatal characteristics such as preterm birth (< 37-week gestational period), neonatal sex (female/male), and breastfeeding within the first 6 weeks (exclusive, complementary, or never).

The general characteristics of the participating women with and without lactational mastitis experience during the postpartum period were compared using the chi-square test. The differences in EPDS and PHQ9 summary scores between the lactational mastitis and non-mastitis groups were compared by the non-parametric test. The differences in the proportions of women with EPDS cut-off scores ≥9 or PHQ9 cut-off scores ≥5 were compared using the chi-square test. The chi-square test was also used to compare the differences in the proportions of women with self-harm or suicidal ideation (with self-harm ideation: item 10 > 0, in EPDS; with suicidal ideation: item 9 > 0, in PHQ9) between the two groups.

A multiple linear regression analysis was performed to evaluate the associations of lactational mastitis with EPDS or PHQ9 summary scores. A multivariable log-binomial regression analysis was conducted to evaluate the relative risk (RR) and 95% confidence interval (CI) for the proportions of women with lactational mastitis presenting with depression symptoms or self-harm/ suicidal ideation. These two regression analyses were adjusted for the following potential confounders: maternal age, education level, employment status, prenatal depression, nulliparity, smoking, alcohol use, maternal comorbidities, delivery modality, preterm birth, neonatal sex, and breastfeeding status.

Stratified analyses were performed to detect whether the associations of lactational mastitis with postpartum depression symptoms differed according to maternal age, employment status, nulliparity, maternal comorbidities, delivery modality, neonatal sex, and breastfeeding status. RRs and 95% CIs were estimated for each subgroup. Finally, sensitivity analysis was performed to address the confounding effects of prenatal depression, smoking/alcohol use, and preterm birth by restricting log-binomial regression analysis to women with no history of prenatal depressive disorder, women with no habit of smoking or alcohol use during pregnancy, or women with full-term birth.

All statistical analyses were performed using R statistical software version 3.2.3 (R Project for Statistical Computing, The R Foundation, Vienna, Austria), and a two-sided value of p of <0.05 was considered statistically significant.

A total of 1,551 pregnant women with singleton pregnancies were included in this study. The demographic characteristics are summarized and compared in Table 1. Of the participating women, the majority were aged <35 years (85.7%), nulliparous (72.4%), had a bachelor’s degree or above (88.7%), employed during pregnancy (79.2%), and had exclusive or complementary breastfeeding status (93.5%). Only a small number of participating women smoked (0.7%) or consumed alcohol (2.1%) during pregnancy, had a history of prenatal depression (0.5%), and delivered prematurely (7.0%). Among the 1,551 participants, 9.5% (N = 147) experienced lactational mastitis during the postpartum period, and the demographic characteristics of these women were not significantly different from those of women who did not experience lactational mastitis.

Among the 1,551 participants, the summary score of the EPDS was 5 (interquartile range [IQR] 3–7), and that of the PHQ9 was 2 (IQR 1–4). The proportion of women with depression symptoms was 28.4% (EPDS ≥9 or PHQ9 ≥ 5). Women who had experienced lactational mastitis during the postpartum period reported higher summary scores of the EPDS (5 [IQR 3–7] vs. 6 [IQR 3–8], p < 0.01) and PHQ9 (2 [IQR 1–4] vs. 3 [IQR 1–5.5], p < 0.01) 6 weeks after delivery than women without experience of lactational mastitis (Table 2). In addition, the proportion of women with depression symptoms was significantly higher in the mastitis group than in the non-mastitis group (38.1% vs. 27.4%, p = 0.008), and similar results were also obtained when these two screening tools were considered independently (EPDS ≥9: 23.8% vs. 16.4%, p = 0.031; PHQ9 ≥ 5: 33.3% vs. 23.9%, p = 0.016) (Table 2). Moreover, women who had experienced lactational mastitis during the postpartum period were found to have higher proportions of self-harm or suicidal ideation (11.6% vs. 6.6%, p = 0.026). When self-harm ideation and suicidal ideation were considered separately, self-harm ideation showed marginal significance, while there was no significant statistical difference in suicidal ideation (with self-harm ideation: 8.2% vs. 4.7%, p = 0.068; with suicidal ideation: 8.2% vs. 5.2%, p = 0.133) (Table 2).

In the multiple linear regression analysis, lactational mastitis was significantly associated with higher EPDS and PHQ9 summary scores during the postpartum period. In the adjusted model, compared with women without lactational mastitis experience, the mean EPDS summary score was higher by 1.14 (β = 1.14, 95% CI: 0.49–1.80, p < 0.01), and the mean PHQ9 summary score was higher by 0.78 (β = 0.78, 95% CI: 0.24–1.33, p < 0.01) in women who had experienced lactational mastitis after delivery (Table 3).

In the multivariable logistic regression analysis, lactational mastitis was significantly associated with an increased risk of postpartum depression (Table 4). Compared with women without lactational mastitis experience, the risk of postpartum depression (EPDS score ≥ 9 or PHQ9 score ≥ 5) increased by 68% (adjusted RR [RR] = 1.68, 95% CI: 1.18–2.4, p < 0.01) in women who had experienced lactational mastitis after delivery (Table 4). The findings were consistent when these two scales were used separately (EPDS score ≥ 9: RR = 1.66, 95% CI: 1.1–2.5, p = 0.02; PHQ9 score ≥ 5: RR = 1.65, 95% CI: 1.14–2.38, p < 0.01) (Table 4). In addition, lactational mastitis was significantly associated with an increased risk of self-harm or suicidal ideation (RR = 1.89, 95% CI: 1.08–3.29). When self-harm ideation and suicidal ideation were considered separately, the associations of lactational mastitis with self-harm ideation showed marginal significantly (RR = 1.84, 95% CI, 0.96–3.52), while there was no significant association with suicidal ideation (RR = 1.69, 95% CI, 0.89–3.23).

We performed a sensitivity analysis to address the impact of confounding effects. By restricting the multiple linear regression analysis to women with prenatal depression, the associations between lactational mastitis and postpartum depression symptoms did not change (Supplementary Table S1). When restricting the analysis to women with smoking/alcohol use habits, the observed associations did not change (Supplementary Table S2). After restricting the analysis to women with preterm births, the associations remained unchanged (Supplementary Table S3).

Further, we performed a stratified analysis by maternal age, employment status, nulliparity, maternal comorbidities, delivery modality, neonatal sex, and breastfeeding status and calculated RRs and 95% CIs for each subgroup (Figure 1). The association between lactational mastitis and postpartum depression symptoms was more robust in participants aged ≥35 years (RR = 2.65, 95% CI: 1.01–6.53, p = 0.03), with maternal comorbidities (RR = 2.63, 95% CI: 1.21–5.7, p = 0.02), and those who delivered a female neonate (RR = 1.91, 95% CI: 1.14–3.2, p = 0.01).