The prevalence of psychiatric symptoms of pregnant and non-pregnant women during the COVID-19 epidemic

The study investigates whether pregnant women experienced different levels of psychiatric symptoms compared to non-pregnant women during the COVID-19 epidemic in China. Given the rapid spread of COVID-19 and documented psychological impacts on the general public, adolescents, and healthcare workers, the authors highlight a gap: limited evidence on mental health among pregnant women during the pandemic. Pregnancy may entail heightened psychological vulnerability due to immune changes, disrupted prenatal care, and concerns about adverse perinatal outcomes seen in SARS and MERS. The purpose is to assess prevalence of depression, anxiety, somatic symptoms, insomnia, and PTSD among pregnant versus non-pregnant women and test whether pregnant women have more severe mental symptoms during COVID-19.

Prior research during COVID-19 indicates substantial psychological impact in the general public (over half reporting moderate-to-severe impact; Wang et al.), adolescents (22% anxiety, 10.4% severe depression; Yang et al.), and healthcare workers (high rates of depression, anxiety, insomnia, and distress; Lai et al.). Historical data from the SARS (2003) outbreak showed high rates of depressive symptoms (31.2%) and PTSD (28.9%) among quarantined individuals (Hawryluck et al.). For pregnant women infected with coronaviruses, studies reported increased risk of adverse outcomes such as preterm birth and neonatal death in SARS and MERS. However, psychological effects specifically among pregnant women during COVID-19 had not been systematically examined. Prior comparative studies outside epidemic contexts are mixed: some report lower depression and PTSD in pregnant women compared with non-pregnant women (Mota et al.; Stampfel et al.), others find similar levels (Leach et al.) or higher anxiety in pregnancy (Adewuya et al.; Uguz et al.). These discrepancies may reflect differences in populations, instruments, criteria, and contextual stressors such as epidemics.

Design: Cross-sectional, anonymous online survey during the COVID-19 epidemic in China. Setting and recruitment: Several Maternal and Child Health Hospitals in Beijing disseminated survey links via social networks and WeChat (Tencent Inc., China). Data collection occurred from February 28 to March 12, 2020, during the period after the epidemic’s peak in China. Inclusion targeted pregnant and non-pregnant women of childbearing age across China; respondents came from 41 cities (about three quarters from Beijing). IRB approval was obtained from the Institute of Psychology, Chinese Academy of Sciences; informed consent was obtained online. Participants: 873 completed questionnaires were received; 14 respondents with a diagnosis of mental illness were excluded. Final analytic sample: 859 women (544 pregnant, 315 non-pregnant); age range 20–47 years; 92.1% Han nationality; 93.2% married. Data collected: Socio-demographics (age, nationality, marital status, occupation, education, height, weight, BMI, residence, smoking, drinking, annual family income, chronic diseases), attention to epidemic information (daily following yes/no), and whether relatives/friends were infected with COVID-19. Psychometric instruments (Chinese versions):

• Depression: PHQ-9, 0–27; cutoff >10 indicates depressive symptoms; reported sensitivity 0.89, specificity 0.97; Cronbach’s alpha 0.87 in this study.
• Anxiety: GAD-7, 0–21; cutoff ≥7 indicates anxiety symptoms; validated among pregnant Chinese women; Cronbach’s alpha 0.93.
• Somatic symptoms: SCL-90 somatization subscale, 12 items, 12–60; cutoff ≥36 indicates significant somatic symptoms; Cronbach’s alpha 0.94.
• Insomnia: ISI, 0–28; cutoff ≥15 indicates insomnia symptoms; Cronbach’s alpha 0.93.
• PTSD: PCL-5, 0–80; cutoff ≥33 indicates PTSD; Cronbach’s alpha 0.96. Statistical analysis: Continuous variables as mean ± SD (Student’s t-test); non-normally distributed scale scores as median (IQR) compared with Mann–Whitney U test; categorical variables as percentages (chi-square test). Multivariable logistic regression assessed independent association between pregnancy status and each psychiatric outcome, adjusting for covariates significant in univariate analyses (age, marital status, weight, BMI, employment, smoking, drinking, annual family income, chronic diseases). Two-tailed alpha = 0.05. Analyses conducted with SPSS v18.0.

Sample characteristics: Pregnant women were younger, more often married, had higher weight and BMI, higher employment; non-pregnant women more often current smokers and drinkers and had lower family income; pregnant women more often had chronic diseases. No significant group differences in nationality, education, height, or attention to epidemic information. Symptom scores (median, IQR):

• PHQ-9: pregnant 2.0 (0.0–7.0) vs non-pregnant 3.0 (0.0–7.0), p<0.001.
• GAD-7: pregnant 0.0 (0.0–5.0) vs non-pregnant 2.0 (0.0–5.0), p<0.001.
• SCL-90 somatization: both 15.0; p=0.364.
• ISI: pregnant 2.0 (0.0–6.0) vs non-pregnant 3.0 (0.0–7.0), p=0.358.
• PCL-5: pregnant 1.0 (0.0–5.0) vs non-pregnant 5.0 (1.0–14.0), p<0.001. Prevalence (%):
• Pregnant: depression 5.3%, anxiety 6.8%, somatic symptoms 2.4%, insomnia 2.6%, PTSD 0.9%.
• Non-pregnant: depression 17.5%, anxiety 17.5%, somatic symptoms 2.5%, insomnia 5.4%, PTSD 5.7%. Adjusted associations (pregnancy vs non-pregnancy):
• Depression: OR 0.23 (95% CI 0.12–0.45), p<0.001.
• Anxiety: OR 0.26 (95% CI 0.16–0.42), p<0.001.
• Insomnia: OR 0.19 (95% CI 0.06–0.58), p=0.003.
• PTSD: OR 0.15 (95% CI 0.04–0.53), p=0.003.
• Somatic symptoms: OR 0.57 (95% CI 0.17–1.89), p=0.359 (not significant). No significant associations between education, income, or occupation and most psychiatric scores except somatization differences by education/occupation in the combined sample.

This first comparative assessment during the COVID-19 epidemic found that pregnant women reported significantly fewer symptoms of depression, anxiety, insomnia, and PTSD than non-pregnant women, even after adjusting for socio-demographic and health-related factors. The reduced symptom burden among pregnant women contrasts with concerns that pregnancy would heighten psychological vulnerability during a pandemic. Potential explanations include positive selection into pregnancy (better baseline mental health and financial stability), increased family attention and emotional support, greater contact with healthcare providers offering guidance, and biological factors (e.g., altered processing of chemosensory anxiety signals and pregnancy-related hormonal changes potentially mitigating PTSD symptoms). Differences from earlier outbreak findings (e.g., higher depressive/PTSD prevalence during SARS) may relate to higher SARS mortality and timing; this survey was conducted after the peak of COVID-19 in China when control measures, improved information transparency, and rising recovery numbers could have alleviated public distress. Literature comparisons show mixed results between pregnant and non-pregnant women across contexts, suggesting that population characteristics, instruments, diagnostic criteria, and situational stressors (such as epidemic stage or occupational exposure) shape mental health outcomes. The findings suggest that pregnancy may confer relative resilience to certain pandemic-related psychological stressors, though somatic symptom prevalence did not differ.

During the COVID-19 epidemic in China (late February to mid-March 2020), pregnant women exhibited significantly lower prevalence and severity of depression, anxiety, insomnia, and PTSD symptoms compared with non-pregnant women, and pregnancy status independently predicted reduced odds of these conditions. These results highlight potential protective factors associated with pregnancy during public health crises. Future work should use longitudinal designs to track trajectories across gestation and postpartum, clarify biological and psychosocial mechanisms underlying observed resilience, assess the impact of healthcare access and support, and develop tailored interventions for women at higher risk, including non-pregnant women of childbearing age.