This cross-sectional study was conducted in Kaiser Permanente Northern California (KPNC), a large integrated health care delivery system that provides comprehensive medical care to over 4.5 million members. All 15 regional service centers (with 48 associated office facilities) have Obstetrics and Gynecology and Behavioral Medicine/Psychiatry Departments. KPNC members include individuals covered by employer-sponsored insurance plans, the insurance exchange and Medicaid. Coverage is provided for approximately 30% of the northern California population. KPNC members are socio-demographically similar to the population living in the geographic area (17). KPNC maintains comprehensive electronic health records (EHR).

Beginning on June 22, 2020 pregnant KPNC members were invited to participate in a KPNC COVID-19 Pregnancy Study (18). The survey collected information on depression and anxiety symptoms, as well as COVID-19-related health, healthcare and economic factors. Pregnant individuals 18 years of age or older and English-speaking were identified from the EHR and recruited via email. At the initial survey launch, all eligible individuals were contacted via email and invited to complete a brief, web-based survey. Every 2 weeks thereafter, all newly eligible individuals were emailed a study invitation. Individuals who did not respond to the initial email invitation received email reminders 7 and 14 days after the initial emails were sent. Participants who started but did not complete the survey received two email reminders, at 7 and 14 days after starting the survey. While study recruitment is ongoing, this study reports on data collected from 6,628 individuals who were at any stage of pregnancy between 1 January 2020 through 30 September 2020.

The Institutional Review Board of KPNC approved all study procedures and individuals indicated informed consent by completing the survey after reviewing the consent information in the recruitment email (protocol number 1600113-11).

Inverse probability weighting was used to account for survey non-response (21% response rate). Inverse probability weights for survey response were calculated using predicted probabilities from a logistic regression model including variables expected to be associated with survey response (race/ethnicity category, insurance type, parity, history of depression or anxiety diagnosis, trimester of pregnancy at the start of the COVID-19 pandemic), additional variables strongly associated with depression or anxiety symptoms (maternal age), and an interaction term for race/ethnicity category and insurance type. Weights ranged from 1.74 to 40.24 in individuals who completed the survey. For the 14 individuals with weights greater than 20, weights were truncated at 20.

Weighted percentages were calculated for all categorical measures overall and by depression and anxiety severity status. For individuals with missing data for the exposure or covariates we used multiple imputation by chained equations (MICE) (25) with 100 imputations to impute missing values. After imputation, weighted multinomial logistic regression models (using the glogit option in the SAS logistic PROC) were run for each imputed dataset excluding participants with missing outcomes values for the respective models (n = 36 for prenatal depression and n = 44 for prenatal anxiety). Results were combined using Rubin’s rules (26). Crude point and interval estimates of odds ratios of depression (severe, moderate, mild vs. none) and anxiety (severe, moderate, mild vs. none) associated with each COVID-19-related factor were conducted. Adjusted models included all covariates (maternal age, parity, insurance status, history of a depression or anxiety disorder, race/ethnicity, and trimester at survey completion) and COVID-19-related factors. Analyses were conducted in SAS version 9.4 (SAS Institute INC, Cary NC).

Overall, 36% of the individuals reported depression symptoms (mild: 25%, medium: 8%, severe: 3%; Table 1) and 35% reported anxiety symptoms (mild: 22%, moderate: 8%, and severe 5%; Table 2). A quarter (25%, n = 1,559) of the individuals reported mild to severe symptoms of both depression and anxiety.

Health and healthcare factors: Individuals who reported having COVID-19 during pregnancy were more likely than individuals who did not to report moderate and severe depression (moderate: 16% vs. 8%, severe: 11% vs. 3%) and severe anxiety symptoms (13% vs. 4%) (Tables 1, 2, respectively). Compared to individuals without high-risk employment, individuals with high-risk employment were more likely to have severe depression (8% vs. 3%) and moderate (13% vs. 7%) or severe (8% vs. 3%) anxiety. Compared to individuals who reported mild/no distress due to prenatal care changes, individuals who reported moderate/extreme distress were more likely to report depressive symptoms across all levels of severity (mild: 32% vs. 22%, moderate: 14% vs. 6%, severe: 7% vs. 1%) and anxiety (mild: 29% vs. 19%, moderate: 14% vs. 5%, severe: 9% vs. 1%).

Economic Factors: Although individuals impacted by pandemic-related economic factors (job loss, partner job loss, and childcare) had a higher prevalence of depression and anxiety overall compared to individuals not impacted by these factors there were few differences by severity (Tables 1, 2). However, compared to individuals with no food insecurity, individuals with food insecurity were more likely to have higher depression severity (mild: 32% vs. 23%, moderate: 15% vs. 7%, severe: 10% vs. 2%) and anxiety severity (mild: 29% vs. 20%, moderate: 13% vs. 7%, severe: 11% vs. 2%).

Crude estimates are reported in Supplementary Tables S2A, S2B and multivariable adjusted estimates for the associations of COVID-19-related factors and depression and anxiety are reported in Figure 1A (depression) and Figure 1B (anxiety) and below.

Our finding of the associations of self-reported COVID-19 and employment in a job with high-risk of COVID-19 exposure with prenatal depression and anxiety severity is consistent with other reports in pregnant individuals (32) and other populations (33, 34). In general, people with high-risk employment report more fatigue, health worries and fear (35). These findings support consideration of high-risk employment as a risk factor for mental health disorders in pregnancy, for use in depression/anxiety screening strategies.

We documented escalating severity of depression and anxiety symptoms associated with distress over COVID-19-related changes in prenatal care. These findings are similar to other studies documenting COVID-19-related changes in prenatal care and greater psychological distress (33, 36–38). The COVID-19 pandemic forced rapid implementation of hybrid models of care that were developing prior to the pandemic. Provider and healthcare system recognition of the psychological distress associated with changes to prenatal care should consider this when evaluating which care-delivery modifications to implement as standard prenatal care post pandemic or in future public health emergencies.

Previous research on the association of job loss or loss of source of income due to COVID-19 with prenatal depression and anxiety has reported mixed results (32, 33, 36, 38, 39). Our study suggested a relationship between participant’s job loss and depression, but not anxiety, while partner’s job loss was not related to psychological distress. However, food insecurity was strongly associated with prenatal depression and anxiety. This relationship between food insecurity and prenatal depression has been previously documented in research conducted prior to the pandemic (40, 41), yet the magnitude reported in our study (over 5-fold odds of severe depression associated with food insecurity) exceeds pre-COVID-19 findings (4-fold odds) (40). To our knowledge this is the first time that food insecurity has been studied in relation to prenatal anxiety. The strong relationship between food insecurity and psychological distress has important implications for clinical guidelines and health policy related to pregnancy in general, not just limited to times of public health emergencies. Addressing food insecurity may have a significant impact on the mental health of pregnant individuals, in addition to nutritional health. Given that not all food insecure individuals are eligible for government assistance, research into the effectiveness of non-governmental forms of food assistance, such as food banks and nutritional educational programs, is needed. Healthcare systems can also help by screening pregnant individuals for food insecurity and directing individuals to available resources.

Changes in childcare was also significantly related to increased odds of depression and anxiety. Other studies of pregnant individuals during the COVID-19 pandemic have documented similar findings (32, 38) highlighting the importance of public policy providing monetary support and childcare options especially during times of crises.

The low survey response rate raises concerns about non-response bias. However, our analyses were weighted to account for these differences. It is possible individuals with higher psychological distress were less likely to participate in the study and may explain why the prevalence of depression and anxiety was lower in our sample than has been reported in some, but not all (4), samples of pregnant individuals during the COVID-19 pandemic. However, we found that individuals who completed the survey were slightly more likely to have a history of depression or anxiety than individuals who did not complete the survey. Additionally, the cross-sectional study design limits our ability to establish causation. However, the likelihood of reverse causation is diminished by two points. It is unlikely that depression and anxiety led to any of the COVID-19-related factors with the exception of losing a job. Additionally, the mental health screeners asked about symptoms in the past 2 weeks while the time frame for the COVID-19-related factors reflected anytime since the start of the pandemic. Further, the depression and anxiety outcomes are based on symptoms in the past 2 weeks. If the COVID-19 related exposures illicit acute symptoms and the exposures occurred earlier in pregnancy, our findings may be an underestimate of the true association. Although this study sample represents a racially, ethnically, and socio-economically diverse population representative of Northern California, generalizability to other populations may be limited.

While there have been several studies on the mental health of pregnant individuals during the COVID-19 pandemic, few studies have contextualized the impact of COVID-19-related factors on prenatal depression and anxiety. Additional strengths include the study’s large sample size, diversity across sociodemographic characteristics, stages in pregnancy, and capture of COVID-19-related factors and psychological distress during the first 6 months of the pandemic and height of stay-at-home orders.

This study contributes to the growing body of research documenting the adverse impact of public health emergencies experienced during pregnancy on maternal mental health (42–45). Findings from this study suggest that the COVID-19 pandemic may have severe mental health repercussions for pregnant individuals, an extremely vulnerable population. Several COVID-19-related health, healthcare and economic factors were found to adversely impact the mental health of pregnant individuals. These findings provide valuable information for the rapid development of policy, clinical and healthcare system interventions to promote the mental health of pregnant individuals as the COVID-19 pandemic continues and to effectively prepare for future pandemics. These study findings may also have broad policy and clinical implications outside of the US. The known short- and long-term consequences associated with prenatal psychological distress highlight the need for support services for pregnant individuals experiencing these COVID-19 related stressors. Healthcare systems should consider screening for identified stressors associated with poor mental health when conducting depression and anxiety screening and provide the appropriate interventions. Counseling services, social support and stress-reduction interventions to promote maternal mental health should be implemented. Additionally, continued postpartum monitoring of individuals who had high prenatal depression and anxiety symptoms during the COVID-19 pandemic and their children and appropriate interventions are important for decreasing the risk of the known adverse maternal and child outcomes associated with poor prenatal mental health.