Racial and Ethnic Disparities in Postpartum Depressive Symptoms Before and During the COVID-19 Pandemic

Abstract

Background:

Existing evidence regarding the impact of the COVID-19 pandemic on postpartum mental health is mixed, with disparate studies showing increased, stable, or even reduced risk of postpartum depression (PPD) amid the pandemic. Furthermore, although it is plausible that the pandemic impacted the mental health of mothers from vulnerable and underserved communities differentially, few studies have characterized racial and ethnic differences in the impact of the COVID-19 pandemic on PPD.

Materials and Methods:

Pregnancy Risk Assessment Monitoring System data for 2018–2019 (pre-pandemic period) and 2020 (peri-pandemic period) from 40 sites (n = 110,779, representing 5,485,137 postpartum women) were used to determine whether rates of PPD changed during the first year of the pandemic. Postpartum depressive symptoms were assessed using the Patient Health Questionnaire 2-item. Average marginal effects and contrasts of predictive margins derived from a multivariable logistic regression model were used to compare the risk of PPD before and during the pandemic in the overall sample and across racial/ethnic subgroups.

Results:

Adjusting for sociodemographic and clinical characteristics, we found that the overall risk for PPD remained stable (0.0 percentage points [pp]; 95% confidence interval [CI]: −0.7, 0.6 pp) in the first year of the pandemic. We detected no statistically significant changes in risk for PPD across seven of eight racial/ethnic groups considered; however, the risk of PPD among non-Hispanic Black women fell by 2.0 pp (95% CI: −3.5, −0.4 pp) relative to the pre-pandemic period.

Conclusion:

We identified important subgroup differences in pandemic-related changes in risk for PPD.

Introduction

Postpartum depression (PPD) is the most common morbidity following childbirth.^1,2 The condition, characterized by symptoms of depressed mood, loss of interest or pleasure, and feelings of worthlessness, can arise at any point during the first year following birth.¹ Without intervention, symptoms may persist for years.³ Failure to detect or treat PPD compromises maternal and child well-being; consequences can include relational difficulties, deliberate self-harm, and suicide.⁴

In the United States, postpartum depressive symptoms (PPDS) declined between 2004 and 2018, with a prevalence of 13.2% in 2018.^5,6 However, the COVID-19 pandemic dramatically altered health care delivery, increasing barriers to preventive postpartum health care services,^7,8 and the treatment of maternal mental health disorders.⁹ Pandemic-related changes in day-to-day activities, unemployment, financial difficulties, fears, and social isolation may have further impacted postpartum mental health.¹⁰ A systematic review on PPD during the COVID-19 pandemic reported a prevalence rate range of 7.0–80%.¹¹ Although some studies have reported an increased prevalence of PPDS during the COVID-19 pandemic,^7,12
–14 several studies have found that the risk of PPD was unchanged or reduced relative to the pre-pandemic period.^15

–18 Some have noted the potentially positive effects of expanded availability of virtual mental health services, improved postpartum bonding, and increased duration of leave for mothers postdelivery during the pandemic.^15,17,19

Most prior research on changes in PPDS during the COVID-19 pandemic has been based on analyses of convenience samples,²⁰ clinical data,^7,12,15 or state-level administrative data sources (e.g., vital statistics)¹³ or surveillance data.¹⁴ Evidence based on larger, more diverse samples is needed to increase generalizability and draw conclusions based on subgroup analyses. Given the persistent racial and ethnic disparities in maternal morbidity and mortality in the United States and the disproportionate burden of COVID-19 borne by communities of color, it is plausible that the COVID-19 pandemic impacted the mental health of mothers from vulnerable and underserved communities differentially.^8,21

–25 Yet, to our knowledge, pandemic-related changes in racial and ethnic disparities in PPDS are yet to be explored in a large, diverse sample.

In the present study, we used population-based surveillance data for 40 sites that participate in the Centers for Disease Control and Prevention (CDC) Pregnancy Risk Assessment Monitoring System (PRAMS) surveillance project to (1) determine how the prevalence of PPDS changed during the first year of the pandemic compared to the pre-pandemic period and (2) examine to what extent pandemic-related changes in PPDS varied by race and ethnicity.

Materials and Methods

Data source and study participants

This study utilized repeated cross-sectional data from the 2018–2020 PRAMS, a population-based surveillance system that collects data related to pre-conception, prenatal and postpartum care. The PRAMS study protocol and survey methodology were developed by the CDC. The survey is administered by state and local health departments to women who gave birth within the previous 2–6 months.²⁶ Details on PRAMS survey methodology have been described previously.²⁷

We received data from PRAMS sites and surveys that met the CDC’s 50% response rate threshold. Sites located in U.S. territories (Puerto Rico) were excluded. We further restricted the sample to sites with PRAMS data available for both the pre-pandemic (2018 and/or 2019) and peri-pandemic (2020) periods. Six sites that had data available for the pre-pandemic period only (IN, NC, NY, OK, RI) or the peri-pandemic period only (AZ) were excluded. Our study sample comprised a total of 40 PRAMS sites, including 38 states (AK, AL, AR, CO, CT, DE, FL, GA, HI, IA, IL, KS, KY, LA, MA, MD, ME, MI, MN, MO, MS, MT, ND, NE, NH, NJ, NM, OR, PA, SD, TN, UT, VA, VT, WA, WI, WV, WY), the District of Columbia, and New York City. After excluding sites that did not meet sample inclusion criteria (described earlier) and records with missing data (see Supplementary Table S1 for a summary of variable missingness), we obtained a final sample of 110,779 postnatal women. The University of South Carolina Institutional Review Board determined the study was exempt from review because it involved secondary analysis of publicly available, de-identified data.

Measures

Dependent variable

Self-reported PPDS were assessed via a modified version of the Patient Health Questionnaire 2-item (PHQ-2) included in the PRAMS questionnaire. This screening tool consists of two questions: (1) “Since your new baby was born, how often have you felt down, depressed, or hopeless?” and (2) “Since your new baby was born, how often have you had little interest or little pleasure in doing things you usually enjoyed?”. Response options include “always,” “often,” “sometimes,” “rarely,” and “never.” Respondents who answered “always” or “often” to either question were classified as experiencing PPDS. Respondents who answered “sometimes,” “rarely,” or “never” to both questions were classified as not having PPDS.^28,29

Independent variable

The COVID-19 pandemic period was categorized as pre-pandemic (2018–2019) or peri-pandemic (2020).

Covariates

Covariate selection was guided by Andersen’s behavioral model and the variables available in the 2018–2020 PRAMS surveys and linked birth certificates.³⁰

Race and ethnicity data were obtained from the linked birth certificate. In our analyses, maternal race/ethnicity was categorized as Hispanic, non-Hispanic (NH) American Indian and Alaska Native (AIAN), NH Asian and Native Hawaiian and Pacific Islander (Asian/NHPI), NH Black, NH White (ref), other/mixed, or unknown (Supplementary Table S2). Other covariates derived from the birth certificate included maternal age (<20, 20–29 [ref], or ≥30 years), educational attainment (less than high school, high school graduate, or college graduate [ref]), marital status (married [ref] or unmarried), receipt of Special Supplemental Nutrition Program for Women, Infants and Children (WIC) benefits during pregnancy (yes or no [ref]), smoking during pregnancy (yes or no [ref]), diabetes during pregnancy (yes or no [ref]), hypertension during pregnancy (yes or no [ref]), whether the infant was living at time of report (yes [ref] or no), whether the infant was born with a birth defect (yes or no [ref]), vaginal delivery (yes [ref] or no), preterm birth (yes or no [ref]; preterm birth defined as birth at <37 weeks gestation), and low birth weight (yes or no [ref]; low birth weight defined as birth weight <2,500 g).

Covariates derived from the PRAMS survey included depression during the 3 months before pregnancy (yes or no [ref]), whether the pregnancy was intended (yes [ref] or no; pregnancies were classified as “intended” if wanted “then” or “sooner”^31,32), and current health insurance coverage (yes or no [ref]).

Region of residence (Midwest, Northeast [ref], South, or West census region) was assigned using information on the location of PRAMS sites.

Statistical analysis

Descriptive statistics, including frequencies (unweighted) and percentages (weighted) with 95% confidence intervals (CI), were calculated to describe the study sample. Furthermore, we compared the distribution of maternal and infant characteristics across PPDS status. The Rao-Scott chi-square test was used to determine the statistical significance of bivariate associations. Multivariable logistic regression was used to examine the association between giving birth during the peri-pandemic period and PPDS. The regression model included the covariates described earlier and interactions between race/ethnicity and COVID-19 pandemic period. Results from this nonlinear model were reported as average marginal effects (AME) to understand how a change in a risk factor (e.g., delivering in the peri-pandemic versus the pre-pandemic period) affects the predicted probability of experiencing PPDS, holding other covariates constant.^33
–35 Next, to facilitate interpretation of the interaction effects, for each category of race/ethnicity, we compared the predicted probabilities of experiencing PPDS in the peri-pandemic and pre-pandemic periods.³⁴ The contrast (i.e., the difference in the predicted probability for the peri-pandemic versus the pre-pandemic period) allowed us to assess whether the association between COVID-19 pandemic period and PPDS varied according to race/ethnicity. Analyses were conducted using the survey commands in Stata, version 16 (StataCorp LLC) to apply sampling weights provided by the PRAMS. Estimates were considered statistically significant if 95% CI did not include 0 (equivalent to a two-sided p < 0.05).

Results

Sample characteristics

The study population included 110,779 women (Table 1). Of those, 73,948 gave birth in the pre-pandemic period (2018–2019 PRAMS sample) and 36,831 delivered in the peri-pandemic period (2020 PRAMS sample). The sample was racially and ethnically diverse; 16.7% of sample women were Hispanic, 0.7% NH AIAN, 5.2% NH Asian/NHPI, 15.7% NH Black, 57.8% NH White, and 3.1% other/mixed race. Overall, the highest proportion of sample women were aged ≥30 years (49.2%), high school graduates (51.7%), married (61.9%), and residents of states in the South census region (36.5%). Around 15% of sample women reported having depression in the 3 months before they got pregnant (15.1%). Close to two-thirds did not receive WIC during pregnancy (67.3%), and most reported not smoking during pregnancy (93.4%). Rates of diabetes and hypertension during pregnancy were 7.3% and 10.5%, respectively. Close to two-thirds of sample women had a vaginal delivery (67.1%). Nearly all had current health insurance coverage (90.7%). Among infants, the prevalence of preterm birth, low birth weight, and birth defects was 9.0%, 7.2%, and 0.2%, respectively.

Table 1.

Characteristics of Women With and Without Postpartum Depressive Symptoms; 40 PRAMS Sites, 2018–2020^a

	Total		With postpartum depressive symptoms		Without postpartum depressive symptoms
	N ^b	% (95% CI)^c	N ^b	% (95% CI)^c	N ^b	% (95% CI)^c	p
Total	110,779		16,224	13.4 (13.1–13.7)	94,555	86.6 (86.3–86.9)
COVID-19 pandemic period
Peri-pandemic	36,831	34.2 (34.1–34.4)	5,426	13.4 (12.9–14.0)	31,405	86.6 (86.0–87.1)	0.894
Pre-pandemic	73,948	65.8 (65.6–65.9)	10,798	13.4 (13.0–13.7)	63,150	86.6 (86.3–87.0)
Race/ethnicity
Hispanic	18,195	16.7 (16.4–17.1)	2,272	12.0 (11.3–12.9)	15,923	88.0 (87.1–88.7)	<0.001
Non-Hispanic AIAN	4,582	0.7 (0.6–0.7)	911	19.5 (17.3–21.9)	3,671	80.5 (78.1–82.7)
Non-Hispanic Asian/NHPI	8,048	5.2 (5.0–5.4)	1,424	17.9 (16.6–19.4)	6,624	82.1 (80.6–83.4)
Non-Hispanic Black	20,213	15.7 (15.4–16.0)	3,958	18.1 (17.2–19.0)	16,255	81.9 (81.0–82.8)
Non-Hispanic White	50,879	57.8 (57.4–58.3)	6,287	11.8 (11.4–12.2)	44,592	88.2 (87.8–88.6)
Other/mixed	6,154	3.1 (2.9–3.2)	1,065	17.1 (15.3–19.2)	5,089	82.9 (80.8–84.7)
Unknown	2,708	0.8 (0.7–0.9)	307	13.7 (10.6–17.5)	2,401	86.3 (82.5–89.4)
Age, years
<20	4,659	4.0 (3.8–4.2)	1,122	22.3 (20.4–24.4)	3,537	77.7 (75.6–79.6)	<0.001
20–29	51,150	46.8 (46.3–47.3)	8,633	15.5 (15.0–16.1)	42,517	84.5 (83.9–85.0)
≥30	54,970	49.2 (48.7–49.6)	6,469	10.6 (10.2–11.0)	48,501	89.4 (89.0–89.8)
Educational attainment
Less than high school	12,465	10.7 (10.4–11.0)	2,315	16.8 (15.7–18.0)	10,150	83.2 (82.0–84.3)	<0.001
High school graduate	58,292	51.7 (51.3–52.2)	10,073	16.0 (15.5–16.5)	48,219	84.0 (83.5–84.5)
College graduate	40,022	37.5 (37.1–38.0)	3,836	8.8 (8.4–9.2)	36,186	91.2 (90.8–91.6)
Married
Yes	66,666	61.9 (61.4–62.3)	7,860	10.8 (10.5–11.2)	58,806	89.2 (88.8–89.5)	<0.001
No	44,113	38.1 (37.7–38.6)	8,364	17.5 (16.9–18.1)	35,749	82.5 (81.9–83.1)
WIC during pregnancy
Yes	38,754	32.7 (32.2–33.1)	7,357	17.5 (16.9–18.2)	31,397	82.5 (81.8–83.1)	<0.001
No	72,025	67.3 (66.9–67.8)	8,867	11.4 (11.0–11.7)	63,158	88.6 (88.3–89.0)
Smoked during pregnancy
Yes	8,747	6.6 (6.3–6.8)	2,012	21.8 (20.3–23.3)	6,735	78.2 (76.7–79.7)	<0.001
No	102,032	93.4 (93.2–93.7)	14,212	12.8 (12.5–13.1)	87,820	87.2 (86.9–87.5)
Diabetes during pregnancy
Yes	8,634	7.3 (7.0–7.5)	1,276	13.5 (12.4–14.6)	7,358	86.5 (85.4–87.6)	0.869
No	102,145	92.7 (92.5–93.0)	14,948	13.4 (13.0–13.7)	87,197	86.6 (86.3–87.0)
Hypertension during pregnancy
Yes	13,888	10.5 (10.2–10.8)	2,152	15.4 (14.4–16.5)	11,736	84.6 (83.5–85.6)	<0.001
No	96,891	89.5 (89.2–89.8)	14,072	13.1 (12.8–13.5)	82,819	86.9 (86.5–87.2)
Infant living at time of report
Yes	110,355	99.9 (99.8–99.9)	16,008	13.3 (13.0–13.6)	94,347	86.7 (86.4–87.0)	<0.001
No	424	0.1 (0.1–0.2)	216	46.8 (39.2–54.4)	208	53.2 (45.6–60.8)
Infant born with birth defect
Yes	453	0.2 (0.2-0.3)	97	18.3 (13.4–24.4)	356	81.7 (75.6–86.6)	0.047
No	110,326	99.8 (99.7–99.8)	16,127	13.4 (13.0–13.7)	94,199	86.6 (86.3–87.0)
Vaginal delivery
Yes	72,098	67.1 (66.7–67.6)	10,134	12.9 (12.6–13.3)	61,964	87.1 (86.7–87.4)	<0.001
No	38,681	32.9 (32.4–33.3)	6,090	14.2 (13.7–14.8)	32,591	85.8 (85.2–86.3)
Preterm birth^d
Yes	18,882	9.0 (8.8–9.2)	3,510	17.3 (16.4–18.3)	15,372	82.7 (81.7–83.6)	<0.001
No	91,897	91.0 (90.8–91.2)	12,714	13.0 (12.6–13.3)	79,183	87.0 (86.7–87.4)
Low birth weight^e
Yes	21,373	7.2 (7.1–7.4)	3,990	18.2 (17.3–19.1)	17,383	81.8 (80.9–82.7)	<0.001
No	89,406	92.8 (92.6–92.9)	12,234	13.0 (12.7–13.3)	77,172	87.0 (86.7–87.3)
Depression before pregnancy
Yes	17,983	15.1 (14.8–15.5)	5,682	29.8 (28.7–30.9)	12,301	70.2 (69.1–71.3)	<0.001
No	92,796	84.9 (84.5–85.2)	10,542	10.4 (10.1–10.8)	82,254	89.6 (89.2–89.9)
Intended pregnancy^f
Yes	65,019	59.7 (59.2–60.2)	7,427	10.6 (10.2–11.0)	57,592	89.4 (89.0–89.8)	<0.001
No	45,760	40.3 (39.8–40.8)	8,797	17.5 (16.9–18.1)	36,963	82.5 (81.9–83.1)
Health insurance coverage
Yes	9,328	9.3 (9.0–9.6)	1,424	14.2 (13.1–15.4)	7,904	85.8 (84.6–86.9)	0.125
No	101,451	90.7 (90.5–91.0)	14,800	13.3 (13.0–13.6)	86,651	86.7 (86.4–87.0)
Census region
Midwest	31,539	27.6 (27.4–27.7)	4,530	12.4 (11.9–13.0)	27,009	87.6 (87.0–88.1)	<0.001
Northeast	24,075	21.9 (21.8–22.1)	3,027	11.7 (11.1–12.3)	21,048	88.3 (87.7–88.9)
South	28,618	36.5 (36.3–36.6)	4,905	15.3 (14.7–16.0)	23,713	84.7 (84.0–85.3)
West	26,547	14.1 (14.0–14.1)	3,762	12.8 (12.2–13.4)	22,785	87.2 (86.6–87.8)

The sample was restricted to 40 PRAMS sites with data available for both the pre-pandemic and peri-pandemic periods: AK, AL, AR, CO, CT, DC, DE, FL, GA, HI, IA, IL, KS, KY, LA, MA, MD, ME, MI, MN, MO, MS, MT, ND, NE, NH, NJ, NM, NYC, OR, PA, SD, TN, UT, VA, VT, WA, WI, WV, WY.

Unweighted frequency.

Weighted prevalence and 95% CI.

Preterm birth is defined as birth at <37 weeks gestation.

Low birth weight is defined as birth weight <2,500 g.

Pregnancies were classified as intended when a respondent indicated that she wanted to be pregnant “then” or “sooner.”

AIAN, American Indian and Alaska Native; CI, confidence interval; NHPI, Native Hawaiian and Pacific Islander; PRAMS, Pregnancy Risk Assessment Monitoring System; WIC, women, infants and children nutrition program.

Prevalence of PPDS

The overall prevalence of PPDS was 13.4% (95% CI:13.1–13.7%). Similar prevalence rates were found in the peri-pandemic (13.4%; 95% CI: 12.9–14.0%) and pre-pandemic (13.4%; 95% CI: 13.0–13.7%) periods. Bivariate comparisons revealed significantly higher prevalence of PPDS among mothers who reported depression in the 3 months before pregnancy (29.8%; 95% CI: 28.7–30.9%; p < 0.001), were aged <20 years (22.3%; 95% CI: 20.4–24.4%; p < 0.001), unmarried (17.5%; 95% CI: 16.9–18.1%; p < 0.001), and did not complete high school (16.8%; 95% CI: 15.7–18.0%; p < 0.001). Higher rates of PPDS were also noted among NH AIAN (19.5%; 95% CI: 17.3–21.9%), NH Black (18.1%; 95% CI: 17.2–19.0%), and NH Asian/NHPI (17.9%; 95% CI: 16.6–19.4%; p < 0.001) mothers.

Associations between PPDS, COVID-19 pandemic, and maternal and infant characteristics

The AME presented in Table 2 show how a change in a risk factor affects the predicted probability of experiencing PPDS, holding other covariates constant. After adjusting for maternal and infant characteristics, we found no relationship between the COVID-19 pandemic and PPDS (AME: 0.00 percentage points [pp]; 95% CI: −0.7, 0.8 pp; p = 0.943). Compared with those aged 20–29 years, younger women aged <20 years were more likely to experience PPDS (AME: 3.1 pp; 95% CI: 1.3, 4.8 pp), whereas women aged ≥30 years were less likely to experience PPDS (AME: −2.2 pp; 95% CI: −2.9, −1.5 pp). Compared with NH White women, the predicted probability of experiencing PPDS was significantly higher among NH AIAN (AME: 3.8 pp; 95% CI: 1.8, 5.8 pp) Asian/NHPI (AME: 11.4 pp; 95% CI: 9.6, 13.1 pp), Black (AME: 3.7 pp; 95% CI: 2.7, 4.7 pp), and other/mixed race (AME: 3.9 pp; 95% CI: 2.1, 5.8 pp) women. Compared with women residing in the Northeast census region, women residing in the South (AME: 2.0 pp; 95% CI: 1.1, 3.0 pp) were more likely to experience PPDS.

Table 2.

Associations Between Postpartum Depressive Symptoms, COVID-19 Pandemic, and Maternal and Infant Characteristics; 40 PRAMS Sites, 2018–2020^a,b

	Postpartum depressive symptoms
	AME	95% CI	p
COVID-19 pandemic period
Peri-pandemic	0.000	−0.007 to 0.006	0.943
Pre-pandemic	Ref.
Race/ethnicity
Hispanic	−0.001	−0.011 to 0.008	0.783
Non-Hispanic AIAN	0.038	0.018 to 0.058	<0.001
Non-Hispanic Asian/NHPI	0.114	0.096 to 0.131	<0.001
Non-Hispanic Black	0.037	0.027 to 0.047	<0.001
Non-Hispanic White	Ref.
Other/mixed	0.039	0.021 to 0.058	<0.001
Unknown	0.018	−0.017 to 0.054	0.313
Age, years
<20	0.031	0.013 to 0.048	0.001
20–29	Ref.
≥30	−0.022	−0.029 to −0.015	<0.001
Educational attainment
Less than high school	0.029	0.016 to 0.042	<0.001
High school graduate	0.031	0.023 to 0.039	<0.001
College graduate	Ref.
Married
Yes	Ref.
No	0.004	−0.004 to 0.012	0.361
WIC during pregnancy
Yes	0.012	0.004 to 0.019	0.002
No	Ref.
Smoked during pregnancy
Yes	0.016	0.003 to 0.028	0.012
No	Ref.
Diabetes during pregnancy
Yes	0.001	−0.010 to 0.013	0.832
No	Ref.
Hypertension during pregnancy
Yes	0.007	−0.003 to 0.017	0.189
No	Ref.
Infant living at time of report
Yes	Ref.
No	0.285	0.211 to 0.358	<0.001
Infant born with birth defect
Yes	0.018	−0.032 to 0.068	0.476
No	Ref.
Vaginal delivery
Yes	Ref.
No	0.012	0.005 to 0.019	0.001
Preterm birth^c
Yes	0.011	0.000 to 0.023	0.058
No	Ref.
Low birth weight^d
Yes	0.011	0.000 to 0.023	0.046
No	Ref.
Depression before pregnancy
Yes	0.169	0.157 to 0.181	<0.001
No	Ref.
Intended pregnancy^e
Yes	Ref.
No	0.029	0.022 to 0.036	<0.001
Health insurance coverage
Yes	0.003	−0.009 to 0.015	0.587
No	Ref.
Census region
Midwest	−0.003	−0.012 to 0.005	0.440
Northeast	Ref.
South	0.020	0.011 to 0.030	<0.001
West	0.006	−0.003 to 0.015	0.171

AME derived from logistic regression model that included the covariates listed earlier and interactions between COVID-19 pandemic period and race/ethnicity. Predicted probabilities and contrasts related to the interaction effects are presented in Table 3.

Preterm birth is defined as birth at <37 weeks gestation.

Low birth weight is defined as birth weight <2,500 g.

Pregnancies were classified as intended when a respondent indicated that she wanted to be pregnant “then” or “sooner”.

AIAN, American Indian and Alaska Native; AME, Average Marginal Effect; CI, confidence interval; NHPI, Native Hawaiian and Pacific Islander; PRAMS, Pregnancy Risk Assessment Monitoring System; Ref., Reference Category; WIC, women, infants and children nutrition program.

Table 3.

Associations Between Postpartum Depressive Symptoms and COVID-19 Pandemic by Maternal Race/Ethnicity; 40 PRAMS Sites, 2018–2020^a,b

	Postpartum depressive symptoms
	Predicted probability
	Pre-pandemic	Post-pandemic	Contrast	95% CI	p
Hispanic	0.117	0.125	0.008	−0.009 to 0.025	0.335
Non-Hispanic AIAN	0.160	0.156	−0.003	−0.048 to 0.041	0.878
Non-Hispanic Asian/NHPI	0.241	0.222	−0.018	−0.053 to 0.016	0.296
Non-Hispanic Black	0.165	0.145	−0.020	−0.035 to −0.004	0.012
Non-Hispanic White	0.120	0.123	0.003	−0.006 to 0.012	0.491
Other/mixed	0.153	0.173	0.020	−0.018 to 0.059	0.305
Unknown	0.125	0.167	0.043	−0.037 to 0.122	0.293

Predicted probabilities and contrasts derived from logistic regression model that included interactions between COVID-19 pandemic period and race/ethnicity. Average marginal effects for the remaining covariates are presented in Table 2.

AIAN, American Indian and Alaska Native; CI, confidence interval; NHPI, Native Hawaiian and Pacific Islander; PRAMS, Pregnancy Risk Assessment Monitoring System.

Associations between PPDS and COVID-19 pandemic by maternal race/ethnicity

In Table 3, we present the predicted probabilities of experiencing PPDS in the peri-pandemic and pre-pandemic periods for each category of race/ethnicity. The contrast allows us to assess to what extent the association between COVID-19 pandemic period and PPDS may vary according to maternal race/ethnicity. Among NH Black women, the predicted probability of experiencing PPDS decreased by 2.0 pp in the peri-pandemic period relative to the pre-pandemic period (95% CI: −3.5, −0.4 pp; p = 0.012). We observed a similar decrease in the probability of PPDS among NH Asian/NHPI women (contrast: −1.8 pp; 95% CI: −5.3, 1.6 pp; p = 0.296); however, the contrast was not statistically significant at the 95% level. In other groups, we observed small but statistically insignificant increases in the probability of PPDS in the peri-pandemic period.

Sensitivity Analyses

Study findings were insensitive to the exclusion of records with unknown race/ethnicity (Supplementary Table S3), alternative treatments of births in January–March 2020 (Supplementary Table S4), using different controls for geographic variation in PPDS (Supplementary Table S5), and using different controls for health insurance coverage (Supplementary Table S6). Overall patterns were similar when the sample was further restricted to sites with PRAMS data available for each of the three years in our sample period (Supplementary Table S7). Finally, we split the peri-pandemic period into “early” (January–June 2020) and “late” (July–December 2020) stages in order to explore how changes in PPDS may have evolved over the course of the first year of the pandemic (Supplementary Table S8). Among NH Black women, estimated contrasts were similar across the early and late stages of the peri-pandemic period. We detected no statistically significant changes in risk of PPDS for the remaining categories of maternal race/ethnicity.

Discussion

This study, based on a geographically, racially, and ethnically diverse sample of U.S. women, found no association between giving birth during the COVID-19 pandemic and PPDS for most women after adjusting for maternal and infant characteristics. Although some prior work points to an increased prevalence PPDS during the COVID-19 pandemic,^7,12
–14 our findings are consistent with several studies that suggest that the risk of PPD remained stable or fell during the pandemic.^15

–18 It is possible that, for most women, pandemic-related disruptions to health systems and social life were offset by the benefits of expanded availability of virtual mental health services, improved postpartum bonding, and increased duration of leave for mothers postdelivery during the pandemic.^15,17,19 In an analysis of vital statistics data for births in the central region of New Jersey, McFarland et al. reported higher levels of PPD among women who gave birth in the first month of the pandemic (March 2020).¹³ However, giving birth during the second month of the pandemic was unrelated to PPDS, suggesting that with additional time and information (and reduced uncertainty), women were able to adapt and cope with pandemic stressors.¹³ Federal COVID-19 relief assistance programs may have further contributed to resilience.^36,37

The null association in the overall sample masked important subgroup heterogeneity. Across most of the racial/ethnic groups considered, weakly positive associations between the COVID-19 pandemic period and the likelihood of experiencing PPDS were small in magnitude and statistically insignificant. In contrast, we found that the prevalence of PPDS among NH Black women decreased by 2.0 pp during the peri-pandemic period. A similar decrease was found among NH Asian/NHPI women, though we could not rule out a null association. We had hypothesized that the incidence of PPDS would increase within minoritized populations, similar to patterns observed for other natural disasters.³⁸ Nonetheless, our findings are consistent with prior work examining changes in prenatal mood among low-income pregnant women during the COVID-19 pandemic that found depression symptomatology among minority women declined.¹⁶ This observation may be attributed to higher levels of self-reliance,³⁹ emotion regulation,³⁹ and resilience^40,41 among NH Black women or, potentially, unexpected positive effects of pandemic-related social restrictions.¹⁶ Further research to elucidate the underlying mechanisms will be essential.

PPDS were prevalent before and during the COVID-19 pandemic. The estimated prevalence rates in our sample, 13.4% in both the pre-pandemic and the peri-pandemic period, were higher than those reported in another U.S.-based study (6.5% and 6.9%, respectively).²⁰ The adoption of a validated and more sensitive tool for the detection of PPDS in the CDC PRAMS could account for some of the observed differences. In addition, the earlier study was limited to mothers who downloaded and utilized a mobile health application. Thus, it is possible that mothers with better digital access, more education, and higher socioeconomic status were overrepresented in the sample and contributed to an underestimation of PPDS. In contrast, a prevalence rate of 75% was reported for convenience sample of 262 postpartum women living in the United States during the pandemic.⁴² The authors attributed the high prevalence to the application of a highly sensitive tool in detecting PPD and the fact that about one-third of the women had a prior history of depression. The use of snowball sampling in recruiting participants could have led to a homogenous study population and further contributed to overstatement of the prevalence of PPDS.

Throughout the 2018–2020 study period, PPDS were more prevalent among NH women from minoritized groups than among NH White women. Furthermore, NH Asian/NHPI mothers had the highest probability of experiencing PPDS in the multivariable analysis. Unlike previous studies that noted poorer maternal health outcomes in Hispanic mothers,^8,43 we found no significant difference in the prevalence of PPDS among Hispanic compared with NH White mothers. We are unclear about the possible reasons for this disparity; nonetheless, our findings highlight the need to redouble efforts aimed at reducing the burden of PPD and the ongoing need for interventions to mitigate racial inequities in maternal mental health.

Our study has several limitations that should be noted. First, PPDS are self-reported in the PRAMS, which could lead to response bias. However, the PHQ-2 screening tool has been shown to be valid and reliable.⁴⁴ Second, associations cannot be interpreted causally owing to the pooled cross-sectional nature of the study. The aim of this study was to document changes in PPDS during the first year of the COVID-19 pandemic. Moving forward, our study findings, based on data for women who gave birth in 2020, should be extended by incorporating more recent data from 2021 and beyond. Future studies should explore the roles of prenatal care,^45,46 postpartum care,⁴⁷ social support,⁴⁸ experiences of systemic and interpersonal racism,^29,49,50 and other stressors^51,52 in understanding the potential mechanisms underlying the subgroup differences.

Conclusion

Although this study found no change in PPDS during the COVID-19 pandemic for most women, the null association in the overall sample masked important subgroup differences. Among NH Black women, the probability of PPDS fell by 2.0 percentage points in the peri-pandemic period compared with the pre-pandemic period. Irrespective of the pandemic period, we observed a disproportionate burden of PPDS among NH women who identified as AIAN, Asian/NHPI, Black, or other/mixed race, highlighting an ongoing need for public health strategies aimed at promoting racial/ethnic equity in maternal mental health.

Footnotes

Acknowledgment

The authors would like to acknowledge the PRAMS Working Group and the CDC.

Authors’ Contribution

I.N.O.: Conceptualization; Formal analysis; Data Curation; Writing - Original Draft; Writing - Review & Editing; Visualization. N.L.H.: Conceptualization; Methodology; Validation; Writing - Review & Editing. B.C.: Conceptualization; Methodology; Writing - Review & Editing. P.H.: Conceptualization; Methodology; Writing - Review & Editing.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

Supplementary Material

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

Supplementary Table S4

Supplementary Table S5

Supplementary Table S6

Supplementary Table S7

Supplementary Table S8

References

American College of Obstetricians and Gynecologists Committee on Obstetric Practice. Screening for perinatal depression. Committee opinion no. 630. Obstet Gynecol, 2015 May;125(5):1268–1271; doi: 10.1097/01.AOG.0000465192.34779.dc

Wang

, Liu

, Shuai

, et al. Mapping global prevalence of depression among postpartum women. Transl Psychiatry, 2021; 11(1):543; doi: 10.1038/s41398-021-01663-6

Hutchens

, Kearney

. Risk factors for postpartum depression: An umbrella review. J Midwifery Womens Health, 2020; 65(1):96–108; doi: 10.1111/jmwh.13067

Slomian

, Honvo

, Emonts

, et al. Consequences of maternal postpartum depression: A systematic review of maternal and infant outcomes. Women’s Health, 2019; 15:1–55; doi: 10.1177/1745506519844044

, Rockhill

, Tong

, et al. Trends in postpartum depressive symptoms—27 states, 2004, 2008, and 2012. MMWR Morb Mortal Wkly Rep, 2017; 66(6):153–158; doi: 10.15585/mmwr.mm6606a1

Bauman

, Ko

, Cox

, et al. Vital signs: Postpartum depressive symptoms and provider discussions about perinatal depression—United States, 2018. MMWR Morb Mortal Wkly Rep, 2020; 69(19):575–581; doi: 10.15585/mmwr.mm6919a2

Sakowicz

, Matovina

, Imeroni

, et al. The association between the COVID-19 pandemic and postpartum care provision. Am J Obstet Gynecol MFM, 2021 Nov;3(6):100460; doi: 10.1016/j.ajogmf.2021.100460

, Hung

, Li

, et al. Racial and ethnic disparities in postpartum care in the Greater Boston area during the COVID-19 Pandemic. JAMA Netw Open, 2022; 5(6):e2216355; doi: 10.1001/jamanetworkopen.2022.16355

DiGregory

, Githere

, Crites

, et al. The impact of COVID-19 on postpartum depression and the responsibility of the healthcare system. Cureus, 2022; 14(8):e27805; doi: 10.7759/cureus.27805

10.

Pietrabissa

, Simpson

. Psychological consequences of social isolation during COVID-19 outbreak. Front Psychol, 2020; 11:2201; doi: 10.3389/fpsyg.2020.02201

11.

Usmani

, Greca

, Javed

, et al. Risk factors for postpartum depression during COVID-19 pandemic: A systematic literature review. J Prim Care Community Health, 2021; 12:21501327211059348; doi: 10.1177/21501327211059348

12.

Rokicki

, Steenland

, Geiger

, et al. Trends in postpartum mental health care before and during COVID‐19. Health Serv Res, 2022; 57(6):1342–1347; doi: 10.1111/1475-6773.14051

13.

McFarland

, McFarland

, Hill

, et al. Postpartum depressive symptoms during the beginning of the COVID-19 pandemic: An examination of population birth data from central New Jersey. Matern Child Health J, 2021; 25(3):353–359; doi: 10.1007/s10995-020-03116-w

14.

Valcarce

, Myrer

, Garces

. Comparison of anxiety and depression among women who gave birth in Utah 2016–2020 using the pregnancy risk assessment monitory system (PRAMS). Utah Women’s Health Review, 2022; doi: 10.26054/0d-46dz-sr1a

15.

Waschmann

, Rosen

, Gievers

, et al. Evaluating the impact of the COVID-19 pandemic on postpartum depression. J Womens Health (Larchmt), 2022; 31(6):772–778; doi: 10.1089/jwh.2021.0428

16.

Silverman

, Medeiros

, Burgos

. Early pregnancy mood before and during COVID-19 community restrictions among women of low socioeconomic status in New York City: A preliminary study. Arch Womens Ment Health, 2020; 23(6):779–782; doi: 10.1007/s00737-020-01061-9

17.

Pariente

, Wissotzky Broder

, Sheiner

, et al. Risk for probable post-partum depression among women during the COVID-19 pandemic. Arch Womens Ment Health, 2020; 23(6):767–773; doi: 10.1007/s00737-020-01075-3

18.

Boekhorst

MGBM

, Muskens

, Hulsbosch

, et al. The COVID-19 outbreak increases maternal stress during pregnancy, but not the risk for postpartum depression. Arch Womens Ment Health, 2021; 24(6):1037–1043; doi: 10.1007/s00737-021-01104-9

19.

Choi

, Records

, Low

, et al. Promotion of maternal–infant mental health and trauma-informed care during the COVID-19 pandemic. J Obstet Gynecol Neonatal Nurs, 2020; 49(5):409–415; doi: 10.1016/j.jogn.2020.07.004

20.

Bajaj

, Salimgaraev

, Zhaunova

, et al. Rates of self-reported postpartum depressive symptoms in the United States before and after the start of the COVID-19 pandemic. J Psychiatr Res, 2022; 151:108–112; doi: 10.1016/j.jpsychires.2022.04.011

21.

Dubowitz

, Dastidar

, Troxel

, et al. Food insecurity in a low-income, predominantly African American cohort following the COVID-19 pandemic. Am J Public Health, 2021; 111(3):494–497; doi: 10.2105/AJPH.2020.306041

22.

Samuel

, Gaskin

, Trujillo

, et al. Race, ethnicity, poverty and the social determinants of the coronavirus divide: U.S. county-level disparities and risk factors. BMC Public Health, 2021; 21(1):1250; doi: 10.1186/s12889-021-11205-w

23.

Tai

DBG

, Shah

, Doubeni

, et al. The disproportionate impact of COVID-19 on racial and ethnic minorities in the United States. Clin Infect Dis, 2021; 72(4):703–706; doi: 10.1093/cid/ciaa815

24.

Gauthier

, Smith

, García

, et al. Exacerbating inequalities: Social networks, racial/ethnic disparities, and the COVID-19 pandemic in the United States. J Gerontol B Psychol Sci Soc Sci, 2020; 76(3):e88–e92; doi: 10.1093/geronb/gbaa117

25.

Shippee

, Akosionu

, Ng

, et al. COVID-19 pandemic: Exacerbating racial/ethnic disparities in long-term services and supports. J Aging Soc Policy, 2020; 32(4-5):323–333; doi: 10.1080/08959420.2020.1772004

26.

United States Centre for Disease Control and Prevention (CDC). PRAMS. United States. 2023. [March 26, 2023]. Available from: https://www.cdc.gov/prams/index.htm.

27.

Shulman

, D’Angelo

, Harrison

, et al. The pregnancy risk assessment monitoring system (PRAMS): Overview of design and methodology. Am J Public Health, 2018; 108(10):1305–1313; doi: 10.2105/AJPH.2018.304563

28.

Haight

, Ko

, Yogman

, et al. Postpartum depressive symptoms and screening opportunities at health care encounters. J Womens Health (Larchmt), 2021; 30(5):731–738; doi: 10.1089/jwh.2020.8373

29.

Weeks

, Zapata

, Rohan

, et al. Are experiences of racial discrimination associated with postpartum depressive symptoms? A multistate analysis of pregnancy risk assessment monitoring system data. J Womens Health (Larchmt), 2022; 31(2):158–166; doi: 10.1089/jwh.2021.0426

30.

Bina

. Predictors of postpartum depression service use: A theory-informed, integrative systematic review. Women Birth, 2020; 33(1):e24–e32; doi: 10.1016/j.wombi.2019.01.006

31.

Finer

, Kost

. Unintended pregnancy rates at the state level. Perspect Sex Reprod Health, 2011; 43(2):78–87; doi: 10.1363/4307811

32.

Keeton

, Hayward

. Pregnancy intention and birth outcomes: Does the relationship differ by age or race? J Womens Health (Larchmt), 2007; 16(4):510–516; doi: 10.1089/jwh.2006.M710

33.

Norton

, Dowd

, Maciejewski

. Marginal effects-quantifying the effect of changes in risk factors in logistic regression models. Jama, 2019; 321(13):1304–1305; doi: 10.1001/jama.2019.1954

34.

Karaca-Mandic

, Norton

, Dowd

. Interaction terms in nonlinear models. Health Serv Res, 2012; 47(1 Pt 1):255–274; doi: 10.1111/j.1475-6773.2011.01314.x

35.

Norton

, Dowd

. Log odds and the interpretation of logit models. Health Serv Res, 2018; 53(2):859–878; doi: 10.1111/1475-6773.12712

36.

Recovery & Economic Support | FEMA.gov. 2023. Available from: https://www.fema.gov/disaster/historic/coronavirus/economic [Jun 30, 2023].

37.

, Luo

, Mu

, et al. Effects of sources of social support and resilience on the mental health of different age groups during the COVID-19 pandemic. BMC Psychiatry, 2021; 21(1):16; doi: 10.1186/s12888-020-03012-1

38.

Harville

, Xiong

, Pridjian

, et al. Postpartum mental health after Hurricane Katrina: A cohort study. BMC Pregnancy Childbirth, 2009; 9(1):21; doi: 10.1186/1471-2393-9-21

39.

Gur

, White

, Waller

, et al. The disproportionate burden of the COVID-19 pandemic among pregnant black women. Psychiatry Res, 2020; 293:113475; doi: 10.1016/j.psychres.2020.113475

40.

Liu

, Hung

, Alberg

, et al. Mental health among pregnant women with COVID-19–related stressors and worries in the United States. Birth, 2021; 48(4):470–479; doi: 10.1111/birt.12554

41.

Assari

, Lankarani

. Association between stressful life events and depression; intersection of race and gender. J Racial Ethn Health Disparities, 2016 Jun;3(2):349–356; doi: 10.1007/s40615-015-0160-5

42.

Goyal

, Beck

, Webb

, et al. Postpartum depressive symptoms and experiences during COVID-19. MCN Am J Matern Child Nurs, 2022; 47(2):77–84; doi: 10.1097/NMC.0000000000000802

43.

Hung

, Liu

, Norregaard

, et al. Analysis of residential segregation and racial and ethnic disparities in severe maternal morbidity before and during the COVID-19 pandemic. JAMA Netw Open, 2022; 5(10):e2237711; doi: 10.1001/jamanetworkopen.2022.37711

44.

Kroenke

, Spitzer

, Williams

JBW

. The Patient Health Questionnaire-2: Validity of a two-item depression screener. Med Care, 2003; 41(11):1284–1292; doi: 10.1097/01.MLR.0000093487.78664.3C

45.

Groulx

, Bagshawe

, Giesbrecht

, et al. Prenatal care disruptions and associations with maternal mental health during the COVID-19 pandemic. Front Glob Womens Health, 2021; 2:648428; doi: 10.3389/fgwh.2021.648428

46.

Lee

. Correlates of and disparities in cancellations or delays of prenatal visits during the Covid-19 pandemic: Emphasis on racial/ethnic minorities and persons with low socioeconomic status. J Racial Ethn Health Disparities, 2023:1–14; doi: 10.1007/s40615-023-01632-3

47.

Bellerose

, Steenland

. Association between the coronavirus disease 2019 (COVID-19) pandemic and national disparities in postpartum visit attendance. Obstet Gynecol, 2023; 141(1):170–172; doi: 10.1097/AOG.0000000000005014

48.

North Carolina PRAMS. Prenatal Social Support. 2015. Available from: https://schs.dph.ncdhhs.gov/data/prams/2015/PPSOCSUP2.html [July 12, 2023].

49.

Huang

, Krupenkin

, Rothschild

, et al. The cost of anti-Asian racism during the COVID-19 pandemic. Nat Hum Behav, 2023; 7(5):682–695; doi: 10.1038/s41562-022-01493-6

50.

Njoroge

WFM

, White

, Waller

, et al. Association of COVID-19 and endemic systemic racism with postpartum anxiety and depression among black birthing individuals. JAMA Psychiatry, 2022; 79(6):600–609; doi: 10.1001/jamapsychiatry.2022.0597

51.

Stone

, Diop

, Declercq

, et al. Stressful events during pregnancy and postpartum depressive symptoms. J Womens Health (Larchmt), 2015; 24(5):384–393; doi: 10.1089/jwh.2014.4857

52.

Schweizer

, Andrews

, Grunewald

, et al. Association of antenatal COVID-19-related stress with postpartum maternal mental health and negative affectivity in infants. JAMA Netw Open, 2023; 6(3):e232969; doi: 10.1001/jamanetworkopen.2023.2969

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.24 MB