Prenatal Mental Health and Its Stress-Process Mechanisms During a Pandemic Lockdown: A Moderated Parallel Mediation Model
Article information
Abstract
Objective
Hundreds of countries have implemented lockdown policies to slow the spread of coronavirus disease-2019 (COVID-19), but the impact of these measures on maternal mental health is not well understood.
Methods
This study integrated a stress-process model to examine the pathways from lockdown-related stressors to prenatal psychological outcomes, with COVID-19 coping strategies (COP) and self-efficacy in managing negative affect (NEG) as mediators and lockdown duration, hours on pandemic-related information, and number of pregnancies as moderators. Pregnant women in Shanghai completed the Regulatory Emotional Self-Efficacy Scale, COVID-19 Coping Scale, Depression, Anxiety, and Stress Scale-21. Structural equation modeling (SEM) was used to test and modify the hypothetical model, and moderated mediation and slope analyses were undertaken.
Results
In the final SEM demonstrating satisfactory fit, three stressors—decreased household income, insufficient daily supplies, and acquired infections—showed positive direct relationships with NEG and COP. Acquired infections, NEG, and COP were identified as direct predictors of mental health outcomes. The relationship between these three stressors and mental health was mediated by NEG and COP. Additionally, the number of pregnancies moderated the mediating effect of COP; this effect was more pronounced among first-time pregnant women than those with multiple pregnancies.
Conclusion
This study provides insights into how lockdown measures impact psychological outcomes in pregnant women quarantined at home. Interventions aimed at increasing coping strategies may be more effective for primiparous women during future public health emergencies.
INTRODUCTION
As the coronavirus disease-2019 (COVID-19) enters its fifth year, the number of reported deaths in April 2024 has dropped to its lowest level since the pandemic began, with just under 0.1 million new monthly cases reported globally. Many countries have discontinued COVID-19-specific reporting and integrated it into respiratory disease surveillance [1]. However, we have to admit that the COVID-19 epidemic has dealt a fatal blow to the world. As of April 28, 2024, over 775 million confirmed cases and over 7 million deaths have been reported globally. Over the past several years, more than 100 countries (including the United States, France, Australia, Thailand, and South Africa) have implemented various restrictive measures at different times to slow the spread of COVID-19. Shanghai, a metropolis of around 24 million people, also entered a phased lockdown on March 28, 2022 under a dynamic zero-COVID policy, and completely returned to normal by June 1st, 2022 [2]. During the lockdown, mandatory stay-at-home reduced community transmission and prevented the overburdening of healthcare systems; however, it may also have had long-term social, behavioral, and economic consequences, resulting in psychological distress, particularly for vulnerable populations [3]. World Health Organization warns that COVID-19 remains a major threat and urges Member States to learn from their experience during this epidemic and maintain rather than dismantle their established COVID-19 infrastructure.
Since the worldwide pandemic began, several studies have documented the psychological impact of the pandemic and its lockdown measures [4]. One crucial factor for well-being is the pandemic’s economic impact [5]. Job loss and financial strain have significantly exacerbated the psychological burden experienced during this period [6]. Research has identified three forms of economic hardship—income loss, job loss, and changes in workload—during the COVID-19 lockdown that are correlated with mental health issues [7]. According to an Austrian study [8] based on an online poll that investigated mental health four weeks after the COVID-19 lockdown, young adults (under 35 years), women, those without jobs, and low-income groups were much more burdened. Furthermore, during the stringent lockdown measures, disruptions to the domestic food supply chain and food production have frequently been identified as key factors contributing to psychological stress [9]. Households have reported reductions in both the quantity and quality of their food consumption during the pandemic due to rising retail prices and decreased income [10,11]. Several studies conducted in North America and 12 sub-Saharan countries have shown that household food insecurity caused by the pandemic or self-quarantine is a significant independent risk factor for poor mental well-being, including higher levels of anxiety, depression, and psychiatric morbidity [12,13].
During the pandemic, women were more likely than men to experience anxiety and depression symptoms, and perinatal women were significantly more susceptible to mental illness than the general population [14]; roughly 12% of prenatal women suffered from depression and up to 22% experienced high levels of anxiety in the third trimester [15]. Major stressors (e.g., health crises and natural disasters) may exacerbate prenatal stress and put gestational women at risk [16]. Studies in China found that prenatal mental disorders were significantly higher after the declaration of COVID-19 than before the outbreak [17]. In Spain, Italy, and the Netherlands, the prevalence of depression and anxiety symptoms among maternal women was also higher during lockdown than the pre-lockdown phase [18,19]. The unprecedented lockdown situation characterized by social isolation, restricted freedom, concerns regarding the impact of the pandemic on pregnancy, potential exposure to COVID-19, unemployment, financial instability, and food insecurity can pose significant challenges to maternal psychological well-being [3,20]. A systematic review has identified various factors associated with an increased prevalence of mental health symptoms in pregnant women during this period including COVID-19 infection, colleagues or family members infected, experience of frontline work, close contact with infected patients, high exposure risk, and high concern about epidemics [21].
Several studies have shown that certain protective variables can successfully mitigate the harmful effects of COVID-19 stressors on psychological states. Effective coping, the primary component of an individual’s response to stressful events, can mitigate certain COVID-19 stressors’ influence on psychological outcomes during pregnancy [22]. During the COVID-19 lockdown in northern Italy, the adoption of coping strategies to counteract antenatal anxiety, depression, and obsessive-compulsive symptoms was important [23]. Pregnant women with relatively high coping capacities and perceived self-efficacy adjusted better to environmental changes or existing stressors and had fewer negative emotional reactions [24]. Although the potential stressors associated with mandatory quarantine during the pandemic and the positive emotions (POS) to alleviate prenatal mental disorders have been widely discussed, knowledge regarding how salient pandemic-related factors and the lockdown affect pregnant women psychologically remains fragmented.
In summary, previous research consistently underscores the detrimental effects of food scarcity, exposure risk, and income reduction during lockdown on maternal mental health. Conversely, it also highlights the beneficial roles of coping styles and self-efficacy in alleviating negative emotions. To investigate the interrelationships among lockdown-related stressors, POS, and maternal mental health during this period, we employ Pearlin et al.’s [25] stress-process model (SPM). This model conceptualizes stress as a process comprising three key components: stressors, resources, and stress outcomes. The SPM has been extensively utilized to explore pathways from common stressors—such as unwanted pregnancy, financial difficulties, and food access—to maternal psychological outcomes [26]. It has also been tested for depression within Chinese prenatal and postnatal populations through structural equation modeling (SEM), which analyzes direct, indirect, and total effects of stressors and mediators on depressive symptoms [27].
By integrating Pearlin’s SPM into our study design (Figure 1), we aim to enhance our understanding of potential pathways leading to adverse psychological consequences among pregnant women during lockdown, thereby providing a reference for the government to formulate humanized restrictive measures for pregnant women in the face of an infectious disease pandemic. Additionally, this study will examine how factors such as lockdown duration, hours spent consuming pandemic-related information, and number of pregnancies moderate the relationship between positive resources and psychological outcomes.
METHODS
Participants
During the COVID-19 lockdown, a hospital-based cross-sectional survey was conducted from April 22, 2022, to June 1, 2022, to recruit a convenience sample of maternal home-quarantined participants. Pregnant women attending prenatal care at the International Peace Maternity and Child Health Hospital of the China Welfare Institute (IPMCH), a tertiary health institution in Shanghai, were invited to complete the survey by scanning a QR code at the hospital or clicking a link sent to their cell phones. Chinese women who met the inclusion criteria were 1) pregnant, 2) older than 18 years, 3) in Shanghai, 4) in mandatory quarantine for the previous 60 days under restrictive measures and still currently in isolation, except for medical appointments, and 5) able to read and write in Chinese. Pregnant women with histories of mental illness were excluded from the study. Informed consent was obtained from all the participants. A total of 1,539 anonymous questionnaires were collected, 48 of which were incomplete, yielding a final sample of 1,491 eligible individuals. The study was approved by the Institutional Review Board of International Peace Maternity and Child Health Hospital (2022-0169).
Measures
Stressors during the lockdown
Several stressors during the lockdown were identified in the study to explore how the pandemic has affected pregnant women. 1) Loss of household income (HI): participants were asked to rate the extent to which lockdown measures had adversely affected their HI in the previous month, with responses ranging from “1=severely affected” to “5=not affected at all”; 2) Lack of daily supplies (DS): participants rated their access to DS (e.g., food) in the previous two weeks on a scale ranging from “1=not met daily needs at all” to “5=completely met daily needs”; 3) COVID-19 infection: participants were asked if they/acquaintances/household members had been medically diagnosed with COVID-19 or identified as being in close contact with the disease in the previous month, using dichotomous items (1=yes, 2=no).
Perceived self-efficacy in managing negative emotions
The Regulatory Emotional Self-Efficacy Scale, initially developed by Caprara et al. [28], measures perceived self-efficacy in managing negative affect (NEG) and POS. NEG comprises two negative effects: despondency-distress (DES) and anger-irritation (ANG), each with four items. In 2009, the scale was modified to include compunction in the NEG subscale to align with Chinese cultural context. Participants rated their responses on a 5-point Likert scale ranging from “1=extremely inconsistent” to “5=highly consistent.” [29] A high score indicated strong belief in one’s ability to manage negative emotions effectively. Previous studies have used the NEG subscale to assess maternal mental health and emotion regulation during the COVID-19 lockdown [30]. Therefore, we utilized the Chinese version of the NEG subscale with 11 items to measure pregnant women’s ability in managing negative emotions. The Cronbach’s alpha for the NEG subscale was 0.956.
COVID-19-related coping strategies
The COVID-19 coping strategies (COP) scale was developed by Yıldırım et al. [31] in the context of COVID-19. Participants rated four items (“I believe that I have the ability to cope with COVID-19”) on a 5-point Likert scale (1=strongly disagree to 5=strongly agree), giving a potential range of 4–20. Higher scores indicated higher levels of COVID-19-related coping strategies. The Cronbach’s alpha for the current sample was 0.886, demonstrating adequate internal consistency.
Mental health
The Depression, Anxiety, and Stress Scale-21 (DASS-21) is a 21-item self-reporting questionnaire with three seven-item subscales: depression, anxiety, and stress [32]. This instrument was used among pregnant individuals during the COVID-19 pandemic to assess mental health outcomes (depression, anxiety, and stress symptoms). In this tool, a 4-point Likert scale ranging from “0=did not apply to me at all” to “3=applied to me a lot or most of the time” was used to assess the state of the participants in the three subscales with five severity ratings. Each item score is multiplied by two, and the final score is the sum of the three subscale values [33]. The Cronbach’s alpha for the DASS-21 was 0.939 in our study.
Socio-demographic characteristics
Socio-demographic characteristics in this study, including age, education, occupation, marital status, HI, residency, and COVID-19 immunization were considered covariates. In addition, the number of pregnancies was categorized as 1, 2, or ≥3. The duration of the lockdown was assessed by the question, “How many days has the community you currently live in been put under lockdown by the government?” (<20/20–25/26–30/31–35/>35 days). For pandemic-related information acquisition, participants were asked how much time they spent every day, on average, looking for information about the epidemic on social media over the past two weeks (none/<0.5/0.5–1/>1 hour).
Statistical analyses
Data were analyzed using IBM SPSS (version 23.0 for descriptive statistics and correlation analyses) and Mplus 8.3 (for SEM, mediation, and moderation analyses; Muthén & Muthén). Descriptive statistics (e.g., frequencies, means, one-way analysis of variance, or t-tests) were performed with NEG, COP, and mental health as dependent variables. Correlations between component measures with continuous distributions were assessed using Spearman’s rank correlation coefficient, and Pearson’s chi-squared (χ2) correlation coefficient for categorical variables. SEM with the maximum likelihood method was performed to test the hypothetical SPM, in which demographic characteristics were introduced as covariates. First, a measurement model was tested using confirmatory factor analysis (CFA). The direct and indirect effects of the relationships between the independent variables (stressors) and mental health were tested using the bias-corrected (BC) bootstrap method. Bootstrapping was set to 5,000, and a 95% BC confidence interval (CI) was reported; if zero was not included in the result, a significant mediating (indirect) effect existed. Subsequently, the hypothesized moderators were incorporated, and the moderated mediation effect of the final model was tested according to the Mplus code proposed by Stride et al. [34] A comparative fit index (CFI), Tucker-Lewis index (TLI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR) were employed to evaluate the optimum model [35]. The model fit was considered acceptable when the CFI and TLI were above 0.90 and the RMSEA and SRMR were below 0.08 [36].
RESULTS
Demographic characteristics and COVID-19-related stressors
As shown in Table 1, pregnant women who had experienced lockdown for 26–35 days were in their first pregnancy and spent more than one hour per day browsing COVID-19 information had a lower NEG (p<0.05) or worse COP (p<0.01) than other groups. More than 80% of home-quarantined pregnant women reported a loss of their HI, and nearly 90% were concerned about a lack of DS (e.g., food) in the past two weeks due to lockdown. Participants who were significantly affected by decreased HI, insufficient DS, and possible infections among acquaintances had poor self-efficacy (p<0.001), COP (p<0.001), and mental health status (p<0.001).
NEG, COP, and mental health among participants with different characteristics and stressors during the COVID-19 lockdown (N=1,491)
Correlation analyses
COP and NEG had significant positive relationships with three key stressors, including loss of HI, acquaintance infection (AI), and lack of DS (p<0.001), with correlation coefficients ranging from 0.114 to 0.274. Similarly, each of the three stressors revealed a slight negative connection with mental health (r from -0.259 to -0.158, p<0.001). Both NEG and COP were significantly associated with mental health with medium intensities (r=-0.471, p<0.001; r=-0.462, p<0.001, respectively).
Model testing
A CFA revealed that a measurement model for NEG, COP, and mental health was acceptable with all factor loadings above 0.5 (p<0.001). However, two path coefficients (from loss of HI and lack of DS to mental health) in the original hypothetical model (Figure 1) were not significant at the p<0.05 level and were deleted from the SEM. The fit indices for the optimized model (Figure 2) were satisfactory (CFI=0.936, TLI=0.930, RMSEA=0.038, SRMR=0.077). Three stressors were shown to be positively and directly related to NEG and COP. Acquired infections (β=-0.089, p<0.001), NEG (β=-0.353, p<0.001), and COP (β=-0.333, p<0.001) directly predicted the degree of mental health.
Final structural equation model with standardized path coefficients. All solid black lines represent paths of the optimized model with p<0.001, while two dashed paths were deleted from the initial model as there was no statistical significance. The deep green areas are three hypothetical moderators in the moderated mediation model. NEG, self-efficacy in managing negative affect; COP, COVID-19 coping strategies; DES, despondency-distress; ANG, anger-irritation; COM, compunction; DL, duration of lockdown; HIS, hours on COVID-19 information seeking; NP, number of pregnancies seeking.
Mediation analyses
A parallel mediation model was employed to explore the influence of stressors on materthat only NP moderated the mediation of COP nal mental health (Y) during the lockdown through the NEG and COP. After adjustments for covariates, the results (Table 2) showed that the total effect of HI, AI, and DS was negatively correlated with mental health (btotal=-10.776, 95% CIBC: -14.010, -7.793), and NEG and COP significantly contributed to the overall indirect effects (bindirect=-6.598, 95% CIBC: -8.251, -5.097). Specifically, NEG and COP fully mediated the relationship between HI, DS, and mental health, and partially mediated the path between AI and mental health (AI-NEG-Y: bpath3=-1.656, 95% CIBC: -2.521, -0.918; AI-COP-Y: bpath4=-1.897, 95% CIBC: -2.985, -1.067), indicating that increased stress from income, food, and infection during the lockdown could significantly reduce positive resources, leading to mental health problems among pregnant women. In addition, the indirect effect of NEG accounted for 28.8% of the total effect, which was less than that of COP (32.5%); however, the difference was not statistically significant, suggesting that the mediating roles of the two variables in the parallel model were roughly the same.
Testing the parallel mediating effect from three stressors to mental health (Y)
Moderated mediation analyses
After controlling for demographic covariates, we respectively incorporated duration of lockdown (DL), hours on COVID-19 information seeking (HIS), and number of pregnancies seeking (NP) as hypothetical moderators in the tested parallel mediation model (Figure 2). Results of testing the moderated mediating effect of stressors on mental health shows that only NP moderated the mediation of COP (bNP*COP=1.501, p<0.05), while NP, DL, and HIS did not moderate the mediation of NEG. The interaction between NEG and the three hypothetical moderators did not yield any significant effect on mental health outcomes. Consequently, DL, HIS, and the moderating effect of NP on the path from NEG to mental health were excluded. In the modified moderated mediation model, the interaction between COP and NP significantly influenced mental health (bNP*COP=1.483, p<0.05).
The results in Table 3 show that NP negatively moderated the mediating effect of COP between the three COVID-19-related stressors and maternal mental health. For women in their first pregnancy, the indirect effect of HI, AI, and DS on mental health through COP was stronger than for mothers with two or more pregnancies. Moreover, simple slope analysis revealed that the negative effect observed between COP and mental health was strongest among women in their first pregnancy (bNP=1=-4.542, bNP=2=-3.059, bNP≥3=-1.576, p<0.001).
Conditional indirect effects of stressors on mental health (Y) through COP at different number of pregnancies
DISCUSSION
It is crucial for policymakers to understand how pandemic restrictions affect mothers’ mental health and to provide a reference for designing maternity-friendly lockdown measures in response to future public health emergencies. Since the beginning of the pandemic, a growing body of research has examined the psychological impact of lockdown measures. However, this study is unique in its use of SEM to design and analyze the mechanisms between lockdown-related risk factors and prenatal psychological outcomes. The SPM was employed during the epidemic to examine the mediating role of POS. Our results indicate a significant relationship between the mental health of pregnant women and three stressors (food, infection, and income) during lockdown, which were fully or partially mediated by maternal COP and NEG. In addition, the moderated mediating results prove that COP is a stronger mediator between COVID-19-related stressors and mental health in first-time pregnant women than in women with two or more pregnancies, and interventions aimed at increasing maternal COP to reduce psychological problems during the lockdown will be more effective among primiparous women; this has not been discussed in previous studies.
Previous studies have indicated that many pandemic-related factors have the potential to influence POS. Higher levels of dysfunctional and problem-focused coping were associated with more financial trouble and social isolation during pregnancy [22], while declining self-efficacy was linked to increased perceived stress [37]. During the quarantine period, pregnant women in our study who experienced stressors such as interpersonal contact with sick acquaintances, limited access to food, and a decline in HI demonstrated a reduction in their COP and NEG levels. Furthermore, several studies have underscored the significance of POS as protective factors in the context of the COVID-19 lockdown, providing evidence of a positive association between effective coping strategies and improved psychological well-being during the pandemic across diverse populations [38,39]. Research has also indicated that higher levels of COVID-19 self-efficacy are associated with reduced suicidal ideation among adults in the United States [40]. Consistent with previous studies among college students, young adults, and vulnerable people [41,42], the protective roles of the COP and NEG against stressors and psychological problems were also confirmed in our sample of pregnant women during lockdown. Specifically, COP- and NEG-related interventions were more effective in alleviating mental distress caused by reduced income and food scarcity due to the COVID-19 pandemic than in alleviating adverse psychological outcomes caused by stress from AIs.
The growing number of deaths and newly diagnosed cases reported online were related to maternal depression rates [43], and regular internet use has affected the mental health of employees and students during the COVID-19 pandemic and lockdown [44]. A survey in China indicated that social media exposure to COVID-19 information enhanced levels of depression and anxiety [45]. These conclusions are similar to those of our study: the time spent browsing pandemic-related information is directly related to maternal NEG, COP, and mental health levels but does not play a moderating role in the relationship between POS and mental health. Browsing epidemic information moderately (within half an hour) per day is better for mental health than not browsing it at all or browsing too frequently.
Furthermore, previous studies have found that individual quarantine is associated with increased anxiety and depression, large disparities have been found between the mental health effects of lockdowns of varied durations in Australia, with the effect of long lockdowns being more than three times larger than that of short lockdowns [46]. Long lockdowns are associated with prolonged psychological distress, financial stress, and social restriction [47]. In our study, there were also differences in the levels of COP, NEG, and mental health among pregnant women with different periods of mandatory quarantine; however, days in quarantine did not play a moderating role in the relationship between POS and mental health. For pregnant women who were restricted to a moderate length of time (26–35 days), their emotional management ability and level of mental health were slightly lower than those who were confined for a short duration and quarantined for a long period of time, suggesting that psychological counseling of pregnant women should be given special attention in areas where the lockdown policy was implemented early and for a long period of time.
Limitations
Despite the significance of these findings, this study has several limitations. First, we employed a cross-sectional design to evaluate measurements concurrently without collecting pre-lockdown baseline data for comparative analysis, owing to the abrupt implementation of restrictions. In the future, longitudinal studies should be conducted to establish causal relationships. Second, many other factors, such as the trimester of pregnancy and behavioral characteristics, such as sleep quality, which could potentially influence psychological outcomes, were not considered in this study. Third, online self-assessment and the convenience sample adopted may have reduced the representativeness and reliability of the findings. Further investigation is required to ascertain whether the findings from pregnant women can be generalized to other adults, vulnerable cohorts, and the general population.
Conclusions
Although our cross-sectional study cannot prove causality, our finding that the COVID-19 pandemic and lockdown could adversely affect maternal mental health is consistent with previous findings. Moreover, we believe that the current study extends our comprehension of how lockdown stressors (income, food, and infection) and protective mediating variables (COP and NEG) predict psychological outcomes in home-quarantined pregnant women. The present study is also the first to find a moderating role of the number of pregnancies on the mediating effect of COP between stressors during the lockdown and maternal mental health, providing a reference for designing pregnancy-friendly restrictions in response to future public health emergencies
Notes
Availability of Data and Material
The datasets used and/or analyzed in this study are available from the corresponding author on reasonable request.
Conflicts of Interest
The authors have no potential conflicts of interest to disclose.
Author Contributions
Conceptualization: Zhen Huang, Man Jiang, Shimeng Liu. Data curation: Dongjian Yang. Funding acquisition: Zhen Huang. Investigation: Lei Chen, Nan Tuo. Methodology: Man Jiang, Nan Tuo, Lei Chen. Writing—original draft: Man Jiang, Zhen Huang. Writing—review & editing: Zhen Huang, Shimeng Liu.
Funding Statement
The study was funded by a grant from the China hospital development institute (project number: CHDI-2023-B-20).
Acknowledgements
We acknowledge the medical staff of the International Peace Maternity and Child Health Hospital who participated in this study for their support and assistance with recruiting participants. We thank the participants for their participation in the study during the COVID-19 lockdown.