Study design and setting
This prospective cohort study included women who delivered via Cesarean section at a tertiary hospital in Henan, China, from 9 October 2021 to 17 May 2022. At this hospital, mothers who deliver by Cesarean section are normally discharged on postpartum day five.
Sampling, inclusion and exclusion criteria
The convenience sample method was applied to identify mothers who met the inclusion and exclusion criteria. The inclusion criteria were: (i) age ≥ 20 years and Cesarean delivery of a singleton who survived; (ii) intent to breastfeed and no obvious contraindications to breastfeeding (e.g., hepatitis B and other active infectious disease diseases; HIV, Treponema pallidum infection; having a neonate with galactosemia or Phenylketonuria); (iii) clear consciousness and ability to understand and answer questions independently; and (iv) available for follow-up via telephone or WeChat. Women were excluded from the cohort if they: (i) had a history of breast surgery including excision biopsies, breast enlargement, breast reduction, or any other surgery involving the breast; (ii) were taking medications that may promote or inhibit lactation after delivery; (iii) had any serious perinatal complications (severe pre-eclampsia or pre-eclampsia, or an NYHA (New York Heart Association) Functional Classification of grade 3 or above); or (iv) had an infant with a critical illness such as cardiopulmonary insufficiency.
Sample size determination
For the analyses presented in this article, we calculated the sample size based on the formula n = Z21 − α/2 × pq /d2. According to a study conducted in the USA in 2010, the prevalence of DOLII among Cesarean section mothers (estimated p) was 59.2% . Based on an allowable deviation of 0.05, statistical power of 90%, and two-tailed significance level of 5%, the minimum required sample size was 300 participants. In anticipation of a 15% dropout rate, we aimed to enroll 345 participants. The actual sample size was 468, indicating that our analyses had sufficient statistical power.
Definition of variables
DOLII was defined as if maternal perception of lactation occurred after 72 h postpartum. Maternal perception of the onset of lactation was assessed by asking women to recall the presence of breast fullness, swelling, and leaking of milk from the breast, and to recall when these signs of lactation II first occurred. This method has been demonstrated to be a valid clinical indicator of lactogenesis II with a sensitivity and specificity of 71.4% and 79.3%, respectively .
We selected the variables from potential confounders of the onset of lactogenesis identified in prior studies from other countries [4, 6, 20,21,22,23,24,25]. Demographic, medical, and breastfeeding factors for the mother and infant were included in the current analysis. The demographic factors were maternal age at delivery, nationality, education level, employment status (any job at the time of birth), mean monthly household income per person, and smoking status and alcohol consumption during pregnancy. The maternal medical factors were gravidity, parity, repeat Cesarean section, pre-pregnancy body mass index (BMI), gestational weight gain (GWG), pregnancy complications, pregnancy course attendance, previous insufficient lactation, serum albumin concentration, hemoglobin concentration, infertility treatment (assisted reproductive technology), insulin treatment, antenatal corticosteroid treatment, type of Cesarean section (elective or emergency), pregnancy outcomes (normal or adverse), type of anesthesia (intraspinal or general), blood loss volume during delivery, and intrapartum fluid volume. The infant medical factors were sex, gestational age, birthweight, height, Apgar score, and neonatal intensive care unit (NICU) admission. The breastfeeding factors were the time of the first breastfeeding session, frequency of breastfeeding in the first 48 h postpartum, and use of formula supplementation during the first 72 h postpartum. The “monthly household income per person” was determined by the total monthly income of all members of the household divided by the total family size. Maternal pre-pregnancy BMI was calculated as weight (kg) / height (m2) and used to categorize the participants according to the Chinese adult BMI classification as underweight (< 18.5 kg / m2), normal weight (18.5–23.9 kg / m2), overweight (24.0–27.9 kg / m2), and obese (≥ 28.0 kg / m2) [26, 27]. GWG was calculated as the difference between the pre pregnancy weight and the last weight measurement during pregnancy. GWG was categorized according to the Society of Chinese Nutrition recommendations , which state that the target GWG should be 11.0–16.0 kg for underweight women, 8.0–14.0 kg for those with normal weight, 7.0–11.0 kg for overweight women, and 5.0–9.0 kg for those with obesity. Maternal GWG was then categorized as excessive, adequate, and inadequate. Pregnancy complications of interest included hypertensive disorders of pregnancy, diabetes (including type 1, type 2, and gestational); thyroid disease (including hypothyroidism and hyperthyroidism); and intrahepatic cholestasis. The maternal serum albumin concentrations (g / L) and hemoglobin concentrations (g / L) were obtained from the medical records of the regular antenatal examinations. Based on our experience, the local normal maternal serum albumin concentration during pregnancy ranged from 35 to 55 g / L. Hemoglobin concentrations were adjusted for altitude as recommended by the World Health Organization  to define anemia in pregnant women as a hemoglobin concentration of less than 110 g / L. The hemoglobin concentration of the pregnant women was not adjusted based on their smoking habits because smoking is currently very rare in Chinese women. The prevalence of smoking was 0.2% in the studied population.
Edinburgh Postnatal Depression Scale 
The Edinburgh Postnatal Depression Scale (EPDS) is a previously validated tool to measure postpartum depressive symptoms [30, 31]. The EPDS consists of 10 domains, covering mood, pleasure, self-blame, anxiety, fear, insomnia, coping skills, sadness, crying, and self-harm. The 10 domains are scored by mothers in accordance with the severity of their symptoms as 0 (never), 1 (occasionally), 2 (often), and 3 (always). Hence, the total EPDS score ranges from 0 to 30, with a higher score indicating a higher risk of postpartum depression. As recommended , we used a cut-off EPDS score of 10 to indicate that the mother had postpartum depressive symptoms and was at high risk of postpartum depression (EPDS score ≥ 10).
During the seven-month recruitment period, researchers attempted to contact all women who delivered a live infant via Cesarean section within the first 24 h after the birth. Baseline demographic information was obtained via face-to-face interviews conducted by the main author and a trained research assistant. Starting 24 h after delivery, we asked the mothers if their milk had come in, if they had noticed breast fullness, swelling, or leakage, and recalled when these signs of lactation II first occurred. The timing of the onset of lactogenesis II was recorded to the nearest hour. If the mother had not experienced lactogenesis II within the first 72 h postpartum, the woman was in contact via telephone or WeChat (Tencent Holdings Limited, Shenzhen, China) daily and was followed up to seven days postpartum. Data were collected on breastfeeding practices in the first 24 and 48 h postpartum, including the frequency of breastfeeding and formula supplementation. Participants completed the EPDS survey on the day of hospital discharge. Data regarding maternal and infant clinical characteristics and the date and time of the birth were extracted from the medical records.
We examined the associations between the outcome (DOLII) and a set of independent variables to identify the factors associated with DOLII. All data cleaning and preparation were done using Microsoft Excel, version 14.1.0 (Microsoft®, Redmond, WA, USA). We used R statistical software version 4.1.0 (SSRI Company, Ltd., Tokyo, Japan) for all analyses.
We conducted an initial descriptive analysis of the characteristics of the study cohort. All covariates were assessed for normality. Covariates that did not meet the criterion of normality were analyzed using non-parametric methods. Descriptive statistics were presented as mean ± standard deviation for continuous variables, and as median (25th–75th percentile) for non-normally distributed continuous variables. Categorical variables were presented as frequencies and proportions. The chi-squared test (or Fisher’s exact test when the cell size was < 5) was used to detect differences in proportions of categorical variables. The Wilcoxon rank-sum test was used to detect differences in medians of continuous variables.
We applied univariate logistic regression to estimate the unadjusted association between DOLII and potential explanatory factors. Factors showing an independent association with DOLII (p < 0.25) were considered eligible for inclusion in the initial multivariable logistic regression model. The cut-off P-value was set at a value larger than the level of significance to obtain as many important variables as possible for inclusion in the model. All independent variables in the initial model were examined for collinearity by calculating the variance inflation factor; using a variance inflation factor of ≥ 4 as the threshold for collinearity , there was no evidence of collinearity. Following the construction of the initial models, a backward elimination method was used to remove variables that were not significantly associated with DOLII (p > 0.05). All variables significant at the α = 0.05 significance level were retained in the final model. We presented the results as unadjusted odds ratios and adjusted odds ratios (aORs), with 95% confidence intervals (CI). The level of statistical significance was set at p < 0.05.
This study was approved by the Scientific Research and Clinical Trial Ethics Committee of the First Affiliated Hospital of Zhengzhou University (project identification code: 2022-KY-0104). During the initial interview, the researchers informed the mothers that their participation was voluntary and that they could withdraw at any time without declaring any reason. Written informed consent was obtained from each participant before study participation.