Network analysis and trajectories of maternal emotional symptoms during labor

Abstract

Background

Owing to the prevalence of maternal mental health problems and the negative effects on both the mothers and infants, this research was conducted to analyze symptoms of maternal emotion and its trajectories during childbirth.

Methods

In this prospective study, Women undergoing vaginal delivery in full-term single pregnancies were enrolled. Symptoms of maternal emotion in the process of childbirth through the scale of Profile Mood States (POMS), encompassing seven dimensions: nervousness, anger, fatigue, depression, confusion, vigour, and self-esteem. All participants were evaluated for their mood state three times, during the latency period (T1), the active period (T2), and the immediate postpartum period (T3).

Results

244 pregnant women were included. Before delivery, there were nervousness, confusion, fatigue and other emotions. As the delivery progressed, negative emotional symptoms intensified, particularly nervousness, fatigue, and confusion, accompanied by a marked decline in vigour. Once the delivery was over, the mood gradually calmed down, mainly positive emotions were observed. There was an evident association between vigour and fatigue only at T3. Confusion, depression, fatigue and nervousness constituted the core symptom group during labor, affecting other emotions of pregnant women. All the network diagrams during the three periods were all stable.

Conclusion

Confusion, depression, fatigue and nervousness constituted the core symptom group during labor. Attaching more importance to the effect of social psychological factors on childbirth, conducting maternal positive mental health education, and adopting maternal case management is imperative.

Trial registration

ChiCTR2000038546; 23/09/2020.

Background

Becoming pregnant and delivering, described as a transition phase, is an existential threshold that mothers have to cross [1]. The process of childbirth is both a natural and normal physiological process for women, as well as a multi-dimensional, complex, and unique experience. It is a powerful source of stress, leading to an increased conflict between a woman’s sense of self-worth and the external environment during the birthing process. This heightened sensitivity to the environment results in psychological stress responses such as excitement, nervousness, anxiety, depression, and fear [2]. Research indicated that the emotional and cognitive experiences of women during childbirth have a significant impact on mothers’ physiological and psychological state after birth, as well as the initial mother-infant bonding with the baby [3]. A positive birthing experience improves the sense of well-being for the mother and promotes the mother-baby relationship. Conversely, negative emotions lead to psychological and physical distress, resulting in postpartum depression [4] or post-traumatic stress disorder [5], even postpartum psychosis. About one third of women consider their childbirth experience to be negative or traumatic [6]. What’s more, experiencing mental health difficulties during childbirth can heighten a woman’s susceptibility to postnatal mental illnesses, specifically postpartum depression and post-traumatic stress disorder [7, 8]. Additionally, it can result in extended labor [9], increased blood loss [10, 11], and a compromised quality of life [12].

When it comes to infants, pregnant women’s negative emotions are related to the higher incidence of cesarean childbirth and less successful breastfeeding [13, 14]. These emotions also lead to lower Apgar scores in newborns, thereby increasing the likelihood of experiencing asphyxia [15], and various studies have additionally indicated that a diminished Apgar score within the normal range significantly heightens the risk of developing metabolic acidosis [16], which was reported to be associated with cognitive disorders [17]. Several studies pointed out that maternal abnormal moods impair neonatal cognitive functions and brain volume, causing learning and memory problems. Inversely, improved maternal moods, and thus neonatal improvement contribute to quicker reach of motor development [18] and physical indicators [19]. In the long term, cardiovascular responses among infants were under the impact of maternal moods during pregnancy [20].

Childbirth is a crucial component of the perinatal period, but it has received limited attention in neuropathological research. Although childbirth is a small part of the entire perinatal period, it is a major turning point that potentially alters childbirth outcomes and the postpartum health trajectory of women. Therefore, the childbirth period should not be overlooked in the neuropathological research of perinatal development. The variables related to labor emotion in the existing literature are mainly the psychological ones related to labor expectation, delivery mode, especially vaginal delivery such as emergency cesarean section or forceps [21]. The relationship variables between positive and negative emotions during the childbirth process have not been clearly defined. Psychological stress symptoms such as tension, anxiety, depression, and fear that arise during childbirth do not exist in isolation; the relationships among these dimensions are complex and interconnected. Understanding the interactions between emotional symptoms helps prevent the occurrence of postpartum stress-related symptoms. Network analysis presents the internal characteristics of a system in the form of a network, mainly composed of “nodes” (representing variables) and “edges” (representing information between variables), and reveals important nodes and network features in the symptom network [22, 23]. In a graphical form, it visualizes the importance of each variable and its complex associations within the network from a holistic perspective, aiding in understanding the mechanisms of symptom occurrence and providing breakthroughs for clinical management. Existing pregnancy monitoring schemes give priority to mothers’ physical health, as a result there is little literature documenting emotional changes during childbirth. Considering the prevalence of mood disorders among pregnant women and the adverse impacts on both the mothers and infants, the purpose of this study is to explore the core emotional symptoms during the three stages of childbirth using network analysis, and explore its trajectories symptoms during childbirth to provide insights for improving childbirth outcomes.

Method

Participants

This prospective study was conducted between October 2021 and October 2022 on women who gave birth to newborns vaginally at Affiliated Maternal and Child Health Care Hospital of Nantong University. It was a single-center study with convenience sampling. The protocol was approved by the Ethics Committee of the Affiliated Maternal and Child Health Care Hospital of Nantong University (No: Y 2020010), and registered in the Chinese Clinical Trial Registry (ChiCTR2000038546; 23/09/2020). Women aged between 20 and 35 undergoing vaginal delivery in full-term single pregnancies were included in the study. Exclusion criteria were women with a history of pre-pregnancy mental disorders, hyperthyroidism, or inherited metabolic diseases. Pregnancies with fetal malformation or maternal depression were also excluded. Individuals who transitioned from vaginal delivery to cesarean section were excluded from the data analysis as they did not complete the full labor process.

According to research, approximately 34% of women report that their childbirth experience was traumatic or negative [6]. Using the PASS 15 software, the sample size was calculated based on the incidence rate of negative childbirth experiences (π) as 34%, with a margin of error of 0.2π. A two-tailed test was conducted with a significance level of α = 0.05. The required sample size was found to be 199 participants. Considering a 20% dropout rate, the final sample size was set at 239 participants.

Evaluation index and method

Mood symptoms were measured by the Profile of Mood States (POMS). It was compiled by American scholars McNair et al. and used for measuring the emotional state. In 1992, Grove et al. simplified it into 40 terms, including negative emotions such as nervousness (6 terms), anger (7 terms), fatigue (5 terms), depression (6 terms) and confusion (5 terms), and positive ones including vigour (6 terms) and self-esteem (5 terms). Each item was rated from five levels, with 0 to 4 standing for “not at all” to “extremely”. The sum of negative emotions minus the sum of positive emotions plus a constant of 100 was used to estimate the emotional state. The overall evaluation of emotional state indicated the state of negative emotions, the higher the scores were, there were more confusion, fatigue or imbalance. The reliability of Chinese simplified POMS ranged from 0.62 to 0.82, with an average of 0.71. This profile involved mothers self-rating their emotional states at three stages: the latency period (when the uterine opening is 3 cm, T1), the active period (when the uterine opening is nearly complete, 8–10 cm, T2), and the immediate postpartum period (within 2 h after delivery, T3).

Data were collected by two investigators. General information (age, education level, marriage, parity, gestational age, complications during pregnancy, delivery mode, neonatal asphyxia and transfer to the department of pediatrics) were collected through medical records.

Statistical analyses

The data was entered by two persons, and the consistency check was performed. Network analysis was carried out using R, Version 4.0.3. This research evaluated the relationship between nodes by describing the centrality. Centrality represents how much, strong, and tightly a node is connected to other nodes. Change a node with high centrality affect many other nodes. The index of centrality includes strength, proximity centrality and intermediate centrality. Strength is the sum of the weighted values of all the connections of a node. It is the quantification of the number and strength of node connections. It quantifies how well a node is directly connected to other nodes and is used to measure the importance of nodes in a network. Proximity centrality is the inverse of the average shortest path length between a node and another one, measuring how closely a node is connected to the others. Mediation centrality is the frequency of a node on the shortest path between any two other nodes, measuring whether a node acts as a mediator of other nodes. Network analysis with R has three basic steps, including network construction, centrality analysis and evaluation of network accuracy and stability. R-package is used to fit an undirected weighted network model, and a concise model is constructed by LASSO and EBIC. Qgraph-package is utilized to visualize the network, in which the positive and negative associated lines are distinguished with different colors. The weighted edge reflects the strength of the relationship between nodes with the thickness of the edge, the thicker the edge is, the stronger the relationship is. Then the centrality index is analyzed, and the research shows that strength is the most robust centrality index. Finally, R-package-bootnet is used to check the accuracy and stability of the network model. The correlation stability (CS) coefficient is used for quantifying the stability of the centrality index. CS coefficient ≥ 0.7 indicates that the centrality index is stable. It is recommended that the CS coefficient be no less than 0.25 and preferably higher than 0.5.

Results

Descriptive statistics

A total of 276 women were assessed for eligibility. During follow-up, six refused to participate at T1, eight participants underwent a conversion from vaginal delivery to cesarean section at T2, and three refused to participate at T3. Ultimately, 244 pregnant women completed the three-time emotional assessment and were included in the data analysis. The participant selection process is detailed in Fig. 1.

Fig. 1

Flowchart of participants from eligibility to data analysis

A total of 244 pregnant women were included, with an average age of 27.18 years and an average BMI of 21.3 kg/m², and an average weight gain during pregnancy is 15.08. Among the participants, 69.3% lived in urban areas, 74.59% were nulliparous, 98.6% were married Throughout the entire pregnancy period, 60.66% of pregnant women remained employed, while 39.34% did not work, 50% of pregnant women received health education during pregnancy. Additionally, 30 (12.30%) were with gestational diabetes, 97.95% experienced full-term pregnancies, 96.31% underwent natural vaginal delivery and 204 (83.61%) used medication to relieve labor pain. (Table 1)

Results of network analysis

With the progress of labor, the total score of negative emotions increased from the initial 103.07 to 109.75, and quickly fell back to 93.70 after delivery (Table 2). We performed a stratified analysis based on labor analgesia used to compare maternal labor mood changes at three time points. At Time1, there were significant differences between the groups in anxiety, fatigue, and depression. At Time2, significant differences were observed between the groups in anger. At Time3, there was no significant difference between the two groups (Table S1).

Before delivery, there were nervousness, confusion, fatigue and other emotions. As the delivery progressed, negative emotional symptoms intensified, particularly nervousness, fatigue, and confusion, accompanied by a marked decline in vigour. Once the delivery was over, the mood gradually calmed down, in addition fatigue, nervousness, confusion and other emotions returned to normal levels, mainly positive emotions were observed (Table 3).

Further, the relationship among maternal emotions at each time-point was analyzed through the network. Since the variables in this study were continuous ones, this model adopted the Gaussian graphical model for correlation analysis to observe the relationship among these symptoms. The three network diagrams respectively represented the relationship among the seven dimensions of the mood profile during the three periods. Green showed a positive correlation and red showed a negative one. The presented network diagram has eliminated the non-significant boundary line between symptoms. Vigour and self-esteem were independent of the other five symptoms, despite they were correlated with the other five symptoms. There was an evident association between vigour and fatigue only at T3, although they were related at T1 and T2, but the strength was small (Figs. 2, 3 and 4).

Figure 5 describes the tightness, weight and mediation of symptoms. During this period, a symptom with a larger value was more likely to act as a core symptom and affect other ones. Confusion, depression, fatigue and nervousness in these three periods constituted the core symptom group during labor, which were more likely to have an effect on other emotions of pregnant women.

The boundary difference between each two nodes in every time period was tested, with 0.05 taken as the test value. Black boxes indicated significant differences in the two edge weights of the nodes, or the centrality of each strength (p < 0.05), gray indicated no statistical significance (p > 0.05), and white indicated specific weight value of the boundary line or node. (Figures 6, 7 and 8).

The stability value of the model in the three periods was used to verify the stability of this model. Generally, stability is greater than 0.5, and the minimum is no less than 0.25. Table 3 shows that the network diagrams in the three periods were all stable.

Fig. 2

Network structure of POMS when the uterine opening is 3 cm. Note: Green showed a positive correlation, and the thicker the line, the stronger the correlation

Fig. 3

Network structure of POMS with uterine orifice opening 8–10 cm. Note: Green showed a positive correlation, and the thicker the line, the stronger the correlation

Fig. 4

Network structure of POMS within 2 h postpartum. Note: Green showed a positive correlation and red showed a negative correlation, and the thicker the line, the stronger the correlation

Fig. 5

Node centrality of the POMS networks across three time-points

Fig. 6

Bootstrapped difference Test for Strength and Closeness of POMS symptoms at T1. Note: Black boxes indicated significant differences in the two edge weights of the nodes, or the centrality of each strength (p < 0.05), gray indicated no statistical significance (p > 0.05), and white indicated specific weight value of the boundary line or node

Fig. 7

Bootstrapped difference Test for Strength and Closeness of POMS symptoms at T2. Note: Black boxes indicated significant differences in the two edge weights of the nodes, or the centrality of each strength (p < 0.05), gray indicated no statistical significance (p > 0.05), and white indicated specific weight value of the boundary line or node

Fig. 8

Bootstrapped difference Test for Strength and Closeness of POMS symptoms at T3. Note: Black boxes indicated significant differences in the two edge weights of the nodes, or the centrality of each strength (p < 0.05), gray indicated no statistical significance (p > 0.05), and white indicated specific weight value of the boundary line or node

Table 1 The baseline characteristics of the pregnant women

Table 2 The POMS scores were recorded for three different time periods(mean ± SD)

Table 3 Correlation stability coefficient of three periods

Discussion

We conducted a study on Chinese women giving birth vaginally, examining emotional changes during childbirth. Panic, depression, fatigue, and nervousness were identified as core symptoms. Each stage of childbirth had distinct core symptom groups, showing strong associations between symptoms. The study revealed similar network patterns across the three stages, providing theoretical evidence for maternal emotion characteristics and targeted interventions.

According to network analysis theory, in this study’s network structure, there is a strong link between depression and tension, aligning partially with previous research [24]. Nervousness and depression are recognized as concurrent symptoms closely related in pathophysiology [25]. Nervousness intensifies the body’s mental and physical response to external stimuli. During labor, stress can trigger various physiological reactions, such as increased activity in the sympathetic-adrenal medulla and hypothalamic-pituitary-adrenal cortex systems [26]. Excessive nervousness or depression reduces norepinephrine secretion and causes hormonal changes, leading to muscle tension, fatigue, delayed labor, and dystocia [27, 28]. Additionally, it poses risks to fetal intelligence and newborns’ lives [27, 29]. Studies indicate a weak positive correlation between fear of childbirth and anxiety sensitivity with somatosensory symptoms [30]. Additionally, the link between anxiety and depression, as well as tension, is somewhat inconsistent with previous results. Anxiety occurs when the brain receives external stimuli, prompting it to send commands to initiate the neuroendocrine system and the sympathetic nervous system [31]. When the sympathetic nervous system becomes more active, it is often accompanied by increased secretion of adrenal medulla, leading to the release of adrenaline and noradrenaline, which enhances the excitatory effects of the sympathetic nervous system [32]. This resulted in elevated blood pressure, increased heart rate, and dilated pupils, manifesting as heightened excitement and tension.

Given the cross-sectional nature of this study, it is evident that anxiety, depression, and tension consistently represent core symptom features, even though bridge symptoms did not appear in the network structure. Intervening in comorbid scenarios of psychological stress reactions is a priority to block or reduce mutual transmission of different symptoms, crucial for attenuating psychological stress reactions. Network analysis studies highlight the importance of intervening in psychological distress and enhancing emotional well-being as high-impact targets throughout the disease trajectory [33]. Targeted interventions for emotional symptoms, particularly social and psychological support, may effectively reduce the overall burden of emotional symptoms [34]. Priority consideration should be given to interventions for maternal emotions, with early implementation upon identification of specific emotions in postpartum women. Developing interventions targeting feelings of anxiety, depression, and tension is crucial for alleviating the overall symptom burden for postpartum women. The childbirth experience may increase conflict between a woman’s self-worth and the external environment, leading to increased sensitivity, cognitive biases, and negative emotions, ultimately impacting the perception of stress during childbirth [35]. Additionally, research suggests that depression and anxiety may impact labor pain, highlighting the growing importance of understanding the two-way interaction between pain and emotional health, particularly in obstetrics [36, 37]. The existing literature on anxiety and pain after surgery in non-obstetric surgical subspecialties suggests that there may be a relationship between pain and anxiety [38].

Curzik and Jokic-Begic [39] also studied dimensions of anxiety sensitivity related to perinatal pain. Physical concern about the anxiety dimension was related to fear of the physical symptoms of anxiety because patients perceived symptoms of anxiety as symptoms of physical illness. Physical worry about the anxiety dimension contributes to anxiety sensitivity during childbirth. Physical worry in the anxiety dimension was related to the greatest pain during childbirth (r = 0.292, P < 0.05). This revealed that fear of physical symptoms of anxiety may also influence labor pain because it gave rise to women’s fear of childbirth. Anxiety sensitivity was thought to promote fear responses when mothers were highly stressed during delivery. This may exacerbate concerns about physiological responses during labor, increase fear of labor, and ultimately increase physiological responses and sensory labor pain.

In this study’s network structure, fatigue was prominent in the first and second stages but weaker in the third, aligning with labor progression. Women experience increasing and persistent fatigue as they enter the active stage, resulting in decreased physical and mental capacity. Fatigue peaks when the cervix fully dilates and diminishes towards the end of labor. The study also found a strong independent correlation between self-esteem and energy, positive emotions, with weak negative correlations to other symptoms, suggesting a mutual inhibitory effect between positive and negative emotions. Prioritizing interventions targeting negative emotions, considering their dominance during childbirth, may benefit outcomes. Moreover, the perception of stressors varies by the individual, and postpartum women experience stress responses. Stress is a natural response to adverse situations that disrupt homeostasis, causing physical and emotional changes and varying degrees of adaptation. Midwives should recognize postpartum women’s individual social factors, including life experiences, personalities, and needs, and provide individualized care, emphasizing psychosocial factors and providing mental health education for the benefit of postpartum women.

Limitation

This study is a single-center sampling study with a small sample size, which may impact the stability of the network. Additionally, this is a cross-sectional study, limiting the determination of causal relationships between symptoms. We also only focused on women from one specific cultural background, further research exploring maternal emotions from different cultural contexts is warranted. Another limitation lies in the dependence on self-report measures only, all the psychiatric symptoms were evaluated and reported by the women themselves. Finally, 74% of the participants were primiparous women and 83.61% were in the labor analgesia population, which may affect the generalizability of the results. In the future, we will expand the sample size to explore changes in maternal labor emotions in populations with different characteristics.

Conclusions

We used the childbirth mood scale to conduct a network analysis of emotional symptoms during childbirth. Anxiety, depression, and tension were identified as core symptoms consistently present across all stages of labor. Interventions targeting these core symptoms may have a significant impact on the overall emotional network structure and should be prioritized in clinical care. Midwives and clinicians should consider these emotional symptoms as key areas for intervention, as addressing them can reduce the overall psychological burden on the mother and improve maternal outcomes. Personalized care strategies, including psychosocial support and mental health education, should be tailored to individual differences, ensuring comprehensive and empathetic maternal care. Early identification and targeted interventions for maternal emotional well-being can lead to better labor experiences and long-term psychological health for both the mother and newborns.