The relationship among pregnancy-related anxiety, perceived social support, family function and resilience in Chinese pregnant women: a structural equation modeling analysis

This is a hospital-based cross-sectional survey. The participants were recruited using convenience sampling method from the maternity outpatient departments of the First Affiliated Hospital of Chongqing Medical University in China from December 2021 to April 2022. Three investigators of our authors, fully trained and certified, used a face-to-face survey to collect the data after signed informed consent was obtained. Specific procedure is as follows: When pregnant women established the health records (gestational week ≥ 11) or attended general maternity examination at Obstetric clinic, investigators would review their health records. Pregnant women who were accorded with the eligibility criteria would be invited to participate in this survey. Then participants filled in the anonymous questionnaires by themselves. For those who did not have a good understanding of the questionnaire content, the research personnel would give corresponding and succinct interpretation instead of biased or instructive suggestions. The investigators were requested to check every questionnaire for accuracy and completeness on the spot. An ethical approval was obtained from the First Affiliated Hospital of Chongqing Medical University for conducting this study.

First, this study is to investigate the prevalence of pregnancy-related anxiety. According to the previous literature, the prevalence of pregnancy-related anxiety in Chinese pregnant women was about 30% [16].With a confidence interval of 95% and 4% tolerable margin of error (d = 4%), the calculated sample size is n =\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left(\frac{{Z}_{1-\alpha /2}}{d}\right)}^{2}p(1-p)$$\end{document}Z1-α/2d2p(1-p) = 504. Then, adding a non-response rate of 10%. Thus, the total sample size is 560.

Second, another purpose of our study is to examine the interrelationships among perceived social support, family function, resilience, and PRA via a structural equation modeling analysis. The minimum sample size requirement of the structural equation was 200 [43]. To sum up, the final sample size required in this study is 560.

The target population of this research is the pregnant women who underwent prenatal examination at the maternity outpatient departments. Pregnant women were eligible to participate in the study if they met the following criteria: aged 18 or above, singleton pregnancy, being able to read and write Chinese, and volunteering to take part in the survey. The exclusion criteria comprised: age less than 18 years; with pregnancy complications (e.g. gestational diabetes mellitus, hypertension, pre-eclampsia), as well as diagnosed and/or being treated for anxiety or depression; fetal disorders and abnormality.

Epi Data version 3.1 and SPSS version 25 were used to record and analyze data. Categorical data and continuous data were listed as frequency and percentage, and mean and standard deviation (SD), respectively. Harman's single-factor test was used to detect the common method bias. Variance inflation factor (VIF) and Cook’s distances were used to test the multicollinearity and outliers, respectively. The relationships between variables was established using Pearson correlation analysis. Statistical significance level was considered as a two-tailed P -value of value less than 0.05.

We implemented a structural equation modeling (SEM) with observed variables using the Maximum Likelihood estimation method to examine the theoretical model hypothesis. Bootstrapping method of repeat sampling by 2000 times was applied to calculate the 95% confidence interval (CI) for the indirect effect, and the indirect effect is considered significant if 95% CI without containing zero. SEM was conducted using AMOS 23.0. The SEM model was assessed using the following indexes and recommended limits: χ²/df < 3; tucker lewis index, TLI > 0.90; normed fit index, NFI > 0.90; root mean square error of approximation, RMSEA < 0.08; goodness of fit index, GFI > 0.90; incremental fit index, IFI > 0.90; comparative fit index, CFI > 0.90 [52].

Common method bias (CMB), multicollinearity and outliers should be reviewed before proceeding. A cross-sectional design and self-reported data might lead to CMB [53]. So, we adopted the Harman's single-factor test to detect it, which is widely used in common method bias testing. After the principal component factor analysis, we extracted 8 eigenvalues greater than 1. And the variance explained by the first factor was only 29.47%, which was less than 40% of the critical standard [54]. Thus, no substantial CMB problem exists in the current study. VIF (< 5.0) and Cook’s distances (< 1.0) were computed to identify the potential existence of multicollinearity and outliers, respectively. Results indicated that there were no problem of outliers and serious multicollinearity in this study (VIF 1.401–1.612; Cook’s distances 0.000–0.883).

We investigated the prevalence of PRA in Chinese pregnant women and explored the mediating role of resilience. PRA occurred in 41.4% of pregnancies in our study, which is similar to the reported prevalence in south-west Nigeria (43.5%) [10]. And the level of PRA yielded by us is much higher than that reported in the Netherlands (11%) [13], Tanzania (25%) [8], Iran (21%) [55] and Qatar (26.5%) [25], whilst it was lower than what was reported in India (55.7%) [11]. Such discrepancy in the percentages reported could be due to various measuring tools used, as well as the cultural multiplicity and sociodemographic diversity of the target population. A Chinese study showed that the prevalence of PRA was about 30% at mid-pregnancy or late-pregnancy under the two-child policy in China [16], which is much lower than that reported in the current study. For pregnant women, especially older pregnant women and multiparas, they would have more and multiple worries or pressure, such as own health, work-family conflicts, childrearing pressure and fetal gender under the three-child policy [56], which might lead to an increase of anxiety during pregnancy. Our findings suggested that the occurrence rate of PRA is high among Chinese pregnant women under the three-child policy, and maternal care providers should give full consideration to reduce the prevalence of PRA by taking effective measures.

This study explored the ways by which perceived social support, family function, and resilience affected the PRA of pregnant women. The results indicated that perceived social support could have a direct effect on the PRA, which is in line with the results reported by Naja in Qatar and Brunton in Australia [25, 57]. This finding can be explained in two ways. First, this might be attributed to the ‘buffering model’ whereby social support serves as a buffer between adverse life events or stress and mental health [58, 59]. The rapid growth of the fetus makes the organs of the mother closer to the maximum functional load in the third trimester. Childbirth fear, worries on the fetal health, and physical discomfort would inevitably bring psychological stress on pregnant women [60]. Adequate social support could protect individuals from deleterious effects of the above stressors, and help pregnant women keep positive emotions and reduce anxiety. Second, social support can practice the coping and assessment ability of individuals, reduce the perceived severity of events and thus play an indirect protective role in individuals’ negative emotions [61]. The association of perceived social support and PRA suggested that support from families, friends and other sources during pregnancy is a focus field for healthcare professionals to assess and manage. Pregnant schools in many hospitals, mainly reflected in maintaining health during pregnancy and in disseminating healthcare knowledge, are an major source of social support for pregnant women at present [62]. Maternity care staffs should offer formal support including psychological counselling and professional informational support to the entire family, promote the adaption of pregnancy, and reduce the unnecessary anxiety. Significantly, spirituality and spiritual care play an important part in providing effective emotional support for patients [63]. So, it is also suggested that medical staffs should raise the awareness about spirituality and spiritual care to offer more support for pregnant women.

Furthermore, our results verified that family function also had a direct impact on PRA. The study conducted by He in China is consistent with our result [64]. In a healthy family, family members can detect the psychological and physical changes of pregnant women, and provide timely material and spiritual help when women cope with anxiety or stressors [65]. However, dysfunctional families often have the characteristic of lower warmth, lower expressiveness and lower cohesion, and but also higher rigidity, conflict and affectionless control, which may be the source of anxiety [66]. Families with domestic violence against women is the reason of many mental illnesses and physical injuries [67]. It is recommended for maternity care providers to encourage family members, especially spouses, strengthen the deeper emotional communication with perinatal women to understand their inner needs of family support and then develop a detailed care plan according to the needs. Community staffs can even proceed regular family visit to ensure that family members are implementing the family support plan, identify problems in family support critically and make effective recommendations to family members.

In line with finding from prior studies [25, 57], resilience had a close negative relationship with PRA in our study. Resilience is a dynamic adaptive process in which individuals respond to adverse events eagerly and make efforts to adapt to the new role or environment [68]. Pregnant women with higher resilience would maintain mental well-being by making full use of available resources to accommodate to the significant changes and learn to cope with pregnancy-related stress or depression effectively [37, 69].

Another important new discovery of our research was that resilience mediated the effects of perceived social support and family function on PRA, respectively. Similarly, a Japanese study also found that resilience had a partial mediating effect between family function and psychological health in hemodialysis patients [70]. Besides, a Chinese research showed that resilience also significantly mediated the links between perceived social support and mental health in teachers [71]. Excellent family function and higher perceived social support can buffer the negative impacts of stressors and provide favorable external environmental conditions for the enhancement of resilience [72, 73]. In a word, greater perceived social support and better family function contributed higher resilience, thus generate positive emotions and alleviate PRA accordingly. Thankfully, resilience has become recognized less as an innate characteristic and more as a set of learnable skills [74]. Cognitive behavioral therapy, mindfulness based stress reduction, and enhancing social support and optimism are potential intervention that serve to increase an individual’s resilience at present [75–77].

The present study had three strengths. First, this is the first work to analyze the interrelationships and potential mechanisms of PRA, perceived social support, family function and resilience of pregnant women by utilizing an SEM, which may have important implications for antenatal care providers and researchers alike. Second, the current study investigated the prevalence of PRA in the context of the three-child policy in China, which provided a reference for developing specific interventions to some extent. Third, a high response rate and the accuracy of the information were guaranteed by face-to-face interviews.

However, a few limitations of this study also merit consideration. First, a convenience sampling method was used, and data were collected in single center, which limited the findings generalize to the whole country and may have introduced bias. Second, the cross-sectional design made it difficult to derive causal interpretations of the relationships among these variables and the claim of the mediating effects of resilience to be somewhat questionable. Third, we did not take the confounding factors into consideration in our study, such as epidemic caused by Corona Virus Disease 2019, which appear to be relevant to exacerbation of anxiety or depression in high-risk group such as pregnant women and health care workers [78–80].