Rates of Major Depressive Disorder and Bipolar Disorder in Black and White Postpartum Women

Rates of Major Depressive Disorder and Bipolar Disorder in Black and White Postpartum Women

Major depressive disorder (MDD) and bipolar disorder (BD) affect 20% of individuals in the first year postpartum, contribute to maternal morbidity and mortality, and adversely affect children and families.^1–4 Identifying postpartum depression is a public health goal as outlined by the United States Preventive Services Task Force and the American College of Obstetrics and Gynecology.^5,6

Studies of racial differences in symptomatology and the rates of postpartum mood disorders are limited. The prevalence of postpartum depression in Black women ranges from 6.8% to 52% and 14%–19.3% in White women.^7–11 Some studies suggest that Black women have a higher prevalence of depressive symptoms and suicidal ideation compared to their White counterparts.^12–14 Variability in assessments, such as the use of postpartum-specific vs general depression scales, a focus on symptoms vs diagnosis, and inconsistent use of diagnostic evaluations contribute to contradictions in findings. Data are limited on the prevalence of BD in the perinatal period, and even less is known about the prevalence of BD among Black women.^15–17 The Epidemiologic Catchment Area Study showed that the rates of BD were similar across racial groups (0.4%); however, bias has historically led to misdiagnosing people that are Black with schizophrenia instead of BD, which resulted in a lower prevalence of BD among Black people.^18–20 The difference in the prevalence of BD as determined by standardized interviews among Black and White postpartum women remains underexplored.

In this secondary analysis of a large-scale postpartum depression screening study at a Midwest urban hospital, we determined the rate of diagnosis of MDD and BD among Black and White postpartum women who screened positive on the Edinburgh Postpartum Depression Scale (EPDS). We compared the rate and severity of specific depressive symptoms, comorbid mental disorders, and history of abuse between Black and White participants.

METHODS

Study Design

The original screening study has been described elsewhere.²¹ Briefly, the investigation was conducted in Pittsburgh, Pennsylvania, at an urban maternity hospital in an academic setting. A nurse or social worker visited women in the maternity ward and offered phone screenings at 4–6 weeks after delivery. Women who were non-English speaking, younger than 18 years old, unable to provide consent, or unable to access a telephone were excluded. All birthing people were included although gender identity was not assessed; therefore, we use the term women in this manuscript. If the woman was not reached within 3 days, a postcard encouraging her to contact the team was sent. Women who were not reached by week 6 were removed from the list. Participants who had a positive EPDS score (greater than or equal to 10) were offered a psychiatric assessment within 2 weeks of the screen. Those who declined the home visit were interviewed by telephone. These diagnostic interviews and symptom assessments were conducted by experienced master’s level psychiatric social workers trained to reliability by the PI (K.L.W.). The study was approved by the University of Pittsburgh Institutional Review Board.

Study Variables

The primary independent variable was self identified race (Black or White). Differences in the prevalence of MDD and BD were determined. We acknowledge that race is a social construct that does not infer causation for disparities. Rates of MDD or BD were determined in women who screened positive on the EPDS and completed a Structured Clinical Interview for DSM-IV (SCID).²² The developers of the EPDS recommended a cut point of 10 or higher for settings with the capacity to evaluate women with positive screens, as was the case in the original research.²³

The secondary outcomes were psychiatric comorbidities, depressive symptom severity, and trauma. We included the following variables from the original dataset: (1) SCID-established comorbid psychiatric diagnoses; (2) severity of and specific types of depressive symptoms with the 29-item Structured Interview Guide for the Hamilton Rating Scale for Depression, Atypical Depression Version (SIGH-ADS), which includes the 17- and 21-item versions of the Hamilton Scale as well as an 8-item scale for atypical depressive symptoms²⁴; (3) postpartum function using the Short Form Survey-12 (SF-12)²⁵; and (4) physical and sexual abuse experienced as a child or adult, using a 4-item yes/no questionnaire where the number of positive responses was summed to create a variable from 0 to 4 for analytic purposes.

Diagnostic Assessment

The complete SCID was administered by master’s level clinicians. The SCID interviewers were trained by viewing 8 standard videotaped diagnostic modules, passing a written examination, and completing reliability ratings with a trained interviewer. Every assessment was reviewed with a board-certified psychiatrist for diagnostic confirmation.

Statistical Analysis

Descriptive statistics were reported as counts and percentages for categorical variables and mean plus SD or median interquartile range (IQR) for continuous normal or continuous skewed variables, respectively. Independent sample t-tests were applied to compare the mean differences between Black and White women on SIGH-ADS scores and symptoms. χ² or Fisher exact tests were used when the expected cell count was less than 5 to compare the proportional differences between Black and White for all categorical variables. Due to the large number of statistical tests conducted in this study, we applied multiple comparison adjustments using the Benjamini-Hochberg procedure.²⁶ Adjusted P values less than .05 were considered statistically significant. This is a relatively conservative method to limit the likelihood of type I error rate inflation (false positives) rather than reporting unadjusted 95% CI for each comparison. We intentionally selected this reporting method and focused our discussion around the most significant differential factors after adjustment. All analyses were done using R (version 4.0.3, https://www.Rproject.org/).²⁷

RESULTS

Sample Derivation

Figure 1 depicts the sample derivation from the original dataset. Table 1 depicts a demographic comparison between those who declined interview and those included in the dataset.

Diagnoses

Positive screens were observed in 19.4% of Black compared to 11.5% of White participants (P < .001). The majority of Black (87%) and White (83%) participants who screened positive on the EPDS met the criteria for either MDD or BD. The rate of MDD was 40% (177/441) in Black and 36% (508/1417) in White participants (Figure 1). White participants with MDD were significantly more likely to be highly educated, be married, and have private insurance and fewer children than Black participants with MDD (Table 2). Black participants had higher rates of trauma (62.7% and 49%, respectively, Table 2) and comorbid posttraumatic stress disorder (PTSD) (14.7% vs 7.7%, respectively, Table 3) compared to White participants. Generalized anxiety disorder and eating disorders were more common in White than Black participants with MDD (Table 3).

The rate of BD was 19% (85/441) and 11.5% (163/ 1417) for Black and White participants, respectively (P < .001, Figure 1). White participants with BD were older, completed higher levels of education and more likely to have private insurance, be married, and have fewer children than Black participants (Table 2). In contrast to MDD, there were no significant differences with respect to past trauma or comorbidities among Black and White participants with BD, except for PTSD, which was more common in Black (28.2%) than in White (19.6%) participants (Table 3).

Symptoms

Among those with MDD, the mean total score on the SIGH-ADS scale was 2 points higher in Black compared to White individuals (22.5 and 20.5, respectively) (Table 4). Scores on the 8-item atypical symptoms subscale of the SIGH-ADS did not significantly differ. The following symptoms were rated higher in severity in Black compared to White participants with MDD (all at or less than P = .004): impairment in work/social activities, decreased appetite, early morning awakening, psychomotor retardation, severity of diurnal variation, paranoid ideation, and social withdrawal (Table 4). Additional symptoms were difficulty with sleep onset (P = .049) and carbohydrate craving (P = .02). The SF-12 Mental and Physical Scale scores did not significantly differ between groups (Table 4).

Among those diagnosed with BD, Black and White women had similar total and atypical symptom scores on the SIGH-ADS (Table 4). Significantly higher symptom scores in Black compared to White subjects were observed for social withdrawal (P = .03) and weight gain (P = .047). The SF-12 Mental and Physical Scale scores did not significantly differ (Table 4).

DISCUSSION

We observed a nearly 2-fold rate of positive screens in Black compared to White participants. Among those who screened positive, Black and White women were similarly likely to be diagnosed with MDD, but Black participants were more likely to be diagnosed with BD. Black women diagnosed with MDD or BD were more socioeconomically disadvantaged, had higher levels of clinical symptoms and higher rates of comorbid PTSD than their White counterparts who were more likely to be educated, privately insured, and married, and have higher rates of generalized anxiety disorder (GAD) and eating disorders. Both groups reported a similar quality of life.

Our results are consistent with previous studies showing higher rates of depressive symptoms²⁸, more severe psychological distress, and lower to similar rates of MDD in nonpregnant Black and White populations,²⁹ as well as postpartum Black and White women diagnosed with a clinician-administered assessment.³⁰ Despite having the same rate of postpartum depression and more severe symptoms than White women, Black women are half as likely to receive a diagnosis of MDD during pregnancy which is also consistent with data among the nonpregnant Black US population.³¹

The equivalent rate of MDD between Black and White participants stands in contrast to studies such as the Survey of American Life, which found markedly lower rates of MDD among nonpregnant Black and White Americans (10.4% vs 17.9%).²⁸ Sociodemographic disparities in Black women are associated with increased risk for postpartum mood disorders and may explain the higher rates of perinatal MDD compared to nonpregnant rates among Black women.^32–34

Structured interviews for mood disorders are critical for comparative diagnostic studies to decrease racial bias.³⁰ Our finding that 40% of Black women who had a positive depression screen had a diagnosis of MDD is strengthened by the sample size and use of structured interviews but contrasts with studies that report less or higher rates of MDD in small sample sizes (ranging from 8 to 162 Black participants).^7–9

Higher rates of BD in Black perinatal people, found in this study, differ from reports that Black and White nonperinatal people have a similar prevalence of BD when diagnosed by standardized assessment.²⁸ Because BD is a severe mental illness that increases the risk of poor pregnancy and birth outcomes when under recognized and under-diagnosed, we aim to increase awareness of postpartum BD across racial groups. Black patients with BD have historically been mis- and under diagnosed. Timely and accurate diagnosis of BD is necessary to prevent inappropriate treatment (ie, unopposed antidepressants), and disparities in prevalence may reflect a need for more focused assessment as well as prevention and intervention strategies among Black perinatal patients.¹⁵

Some investigators have suggested that optimal cutoffs for perinatal depressive screening should be lower than the standard cutoffs when assessing Black perinatal individuals of low socioecnomic status.⁸ It is possible that Black participants with negative screens would have been diagnosed with MDD if they were assessed with a diagnostic examination. Few studies have investigated the sensitivity and specificity metrics of the EPDS in the Black population. Without validation, the optimal cut point for a positive EPDS in the Black perinatal population is unclear.

The clinical characteristics that differed between groups are important for intervention planning. We found that a history of trauma as well as comorbid PTSD was common among those diagnosed with a mood disorder.^35,36 The rates of MDD in both groups in our sample are higher than the general US population.³⁷ Black participants with MDD were disproportionately affected by abuse. Women with BD of both races had strikingly similar and high rates of abuse. In an evaluation of the relationship between childhood maltreatment and depressive symptom severity, clinical symptoms of MDD were significantly associated with depression scores and childhood trauma.³⁸ Symptoms associated with maltreatment were anxiety, psychomotor retardation, diurnal variation, and disordered sleep,³⁸ with the latter 3 being included in the set of symptoms we identified as more common in Black compared to White individuals with MDD.

Perhaps most importantly, childhood trauma also was associated with the presentation and severity of BD, including suicidality, level of functioning between episodes, earlier onset of disease, rapid cycling, and comorbid substance abuse.³⁹ An important environmental risk factor for postpartum BD may be trauma history, which warrants consideration of interventions for perinatal women with a trauma history.

Limitations

The groups of Black and White participants in this study cannot be assumed to be generalizable to all populations of perinatal individuals of these races. Although demographic data were included for the participants who underwent the diagnostic interview, the initial study was not designed to measure parameters of social or structural determinants of health. Additionally, screened individuals did not complete diagnostic assessments if they scored less than 10 on the EPDS and/or declined the diagnostic interview.

We acknowledge that using White women as a comparison population can appear to equate “White” with “standard.” Our intent is to demonstrate that certain differences in the presentation of postpartum mood disorders exist between populations and may be useful to further investigate to establish culturally appropriate and strategic interventions.

Self-report measures, and even standardized clinician-administered assessments, are inherently subjective. Though the EPDS is designed to be patient sensitive in phrasing questions and eliciting responses, thresholds for positive responses may differ between individuals and between groups. We also acknowledge that many individuals identify with several or no racial groups. An attempt was made to include LatinX ethnicity as a third group, but the number of participants was insufficient for analysis. Thus, the complexities of racial and ethnic identity may have been lost as it relates to postpartum mental health.

The inclusion criteria for the original screening protocol were broad and inclusive since the objective was to characterize a population of women defined by the delivery of a live infant. This was a population defined by obstetrical delivery and not by any psychiatric variable. No data were collected on whether the participants were receiving psychiatric treatment; however, it is likely that successfully treated women did not screen positive at the initial telephone contact. Some data were not collected or stratified based on treatment status, so we cannot speculate as to the influence of prior treatment on diagnosis or presentation in the postpartum setting.

CONCLUSION

These results compel advocacy for the allocation of resources and services for the prevention and treatment of postpartum mood disorders, with specific attention to Black populations of low socioeconomic status. Such efforts could reduce morbidity and mortality for peripartum women, their children, and their families. Due to the high rates of trauma in this population, investigations to explore the integration of trauma-informed care are needed to improve outcomes. Research strategies to define the rate and characteristics of postpartum major mood disorders in populations matched for sociodemographic characteristics and social determinants of health between Black and White populations would elucidate the drivers of the differences observed here.

Article Information

Published Online: November 20, 2024. https://doi.org/10.4088/JCP.23m15023
© 2024 Physicians Postgraduate Press, Inc.
Submitted: July 17, 2023; accepted August 23, 2024.
To Cite: Burchfield TN, Yang A, Wisner KL, et al. Rates of major depressive disorder and bipolar disorder in Black and White postpartum women. J Clin Psychiatry. 2024;85(4):23m15023.
Author Affiliations: Northwestern University Feinberg School of Medicine, Asher Center for the Study and Treatment of Depressive Disorders, Chicago, Illinois (Burchfield, Yang, Wisner, Clark); University of Toronto, Women’s College Hospital, Toronto, Ontario, Canada (Clark).
Corresponding Author: Crystal T. Clark, MD, MSc, Crystal T. Clark, MD, MSc, Department of Psyhciatry, Women’s College Hospital, 76 Grenville St, 7th Floor Toronto, ON M5S 1B2, Canada (crystal.clark@wchospital.ca).
Relevant Financial Relationships: Dr Clark serves as a mental health consultant for 7 Starling and has served on the advisory boards for Biogen and 6 Sense Strategy group, for which she has received honoraria. Dr Wisner, Mss Burchfield, and Yang have no financial relationships to disclose.
Funding/Support: This research was supported by funding by the National Institute of Child Health and Human Development (R01 MH071825, Identification and Therapy of Postpartum Depression; PI, Wisner) and the Asher Center for the Study and Treatment of Depressive Disorders.
Role of the Sponsor: The sponsors had no role in the design, management, analysis, or interpretation of the data and did not aid in the creation of the manuscript.
Previous Presentation: These data were presented at the International Society for Bipolar Disorders conference; June 23, 2023; Chicago, Illinois.
Acknowledgments: We thank Karlene Cunningham, PhD, at East Carolina University, and Sheehan Fisher, PhD, at Northwestern University, for their contributions to the manuscript regarding inclusive language and reporting. Dr Fisher is on an advisory board for Woebot Health; is a consultant for Mavida Health and Nurtur; had received grant/research support from the National Institute of Mental Health and the National Institute on Minority Health and Health Disparities. Dr Cunningham has no relevant financial relationships to disclose.

Clinical Points

• Little is known about the differences in prevalence, characterization of symptoms, and comorbidities in postpartum major depressive disorder and bipolar disorder in Black and White mothers.
• In this study, Black and White participants did not differ significantly in the rate of major depressive disorder but significantly differed in diagnosis of bipolar disorder (Black 19% and White 11.5%) postpartum.
• The study demonstrates that trauma history has a striking association with bipolar disorder regardless of self identified race.
• The data support advocacy and allocation of funding and resources toward additional services for the prevention and treatment of postpartum mood disorders in high-risk birthing individuals, with specific attention to Black populations of low socioeconomic status.

The post Rates of Major Depressive Disorder and Bipolar Disorder in Black and White Postpartum Women appeared first on Psychiatrist.com.