Number of significant life events and perinatal depression in a diverse rural population: A brief report of a cross-sectional study

Abstract

Background:

Significant life events (SLEs) correlate with perinatal depression (PD) risk; however, little is known about this association in rural populations.

Objectives:

Using the Neuman systems model, we hypothesized that individuals with higher SLEs would have higher PD risk, and we evaluated this association with data collected from six of Utah’s diverse rural health districts.

Design/Methods:

This cross-sectional study included pregnant and postpartum individuals visiting six rural public health clinics who completed screening with the Edinburgh Postnatal Depression Scale (EPDS) and an 8-item SLE assessment in Spanish or English. Multinomial logistic regression assessed the association between SLEs (0, 1, and ⩾2) and PD risk status (EPDS ⩾9).

Results:

Data from 4245 screening surveys identified 2 primary racial groups: white (79.8%) and American Indian/Alaska Native (7.6%), with 20.2% of individuals identifying as Hispanic. Overall, 49.4% of individuals reported one SLE and 10.5% reported two or more SLEs. Significant predictors for PD risk included 1 SLE (OR = 1.84, 95% CI [1.56, 2.15]) or ⩾2 SLEs (5.18, 95% CI [4.31, 6.23]), regardless of racial/ethnic background.

Conclusion:

Screening for PD risk in a rural population should include an assessment of SLEs, given the relationship between the two. Improving culturally appropriate local resources to support individuals and their families during the childbearing year should be considered, particularly for those experiencing SLEs.

Plain language summary

The relationship between significant life events and perinatal depression in non-urban persons with different ethnicities

Why was the study done? Experiencing stressful events can make it harder for a person to cope with internal and external stressors, which can increase the risk of perinatal depression (PD). This connection needs more study, especially in rural areas. This study uses the Neuman Systems Model (NSM), which looks at different kinds of stressors such as those within a person, between people, and external factors that can disrupt balance.

What did the researchers do? The study described the connection between stressful life events (SLEs) and perinatal depression (PD) in a rural population of pregnant and postpartum individuals to test the idea that more SLEs are linked to a higher risk of PD. This study used a cross-sectional design, meaning data was collected at one point in time. The study team looked at pregnant and postpartum individuals visiting six rural public health clinics in remote areas of Utah. Individuals completed a survey that included questions about their demographics, the Edinburgh Postnatal Depression Scale (EPDS) assessing mental health, and eight questions about significant life events (SLEs) in the past 12 months. SLEs included events like job loss, trouble paying bills, relationship problems, abuse, and moving. The team looked at relationships between SLEs and the risk of PD, considering factors like race, ethnicity, age, and insurance type.

What did the researchers find? Among the 4245 individuals, common SLEs reported were trouble paying bills (24.1%) and moving (25.9%). 26.3% of individuals had a positive screen for PD (EPDS >=9). Individuals with one SLE had almost a two-fold increase in PD risk, while those with two or more SLEs had the greatest risk. Specifically, the adjusted odds ratio for one SLE was 1.84, and for two or more SLEs was 5.18.

What does this mean? In this western rural region, over 26% showed a positive screen for perinatal depression. Experiencing multiple SLEs increased the risk for PD. Screening for SLEs should be part of routine screening to improve mental health among rural childbearing persons.

Background

Environmental stressors represented by past-year significant life events (SLEs) can impact an individual’s ability to maintain mental wellness and thus correlate with increased risk for perinatal depression (PD). However, this relationship is understudied in rural populations, including Latinos and Native Americans. Screening for SLEs is often conducted as part of the Pregnancy Risk Monitoring System (PRAMS) in most states^1,2 with documented associations between SLEs and PD. ³

SLEs are defined as life events occurring in the most recent 12 months and are known to impact health outcomes. According to the Neuman systems model (NSM), intrapersonal stressors (e.g., lifetime history of sexual abuse and partner not wanting pregnancy), interpersonal stressors (e.g., arguing and separation from spouse), and extra-personal stressors (e.g., job loss despite wanting to work, partner job loss, inability to pay bills, and relocation) disrupt balance. ⁴ The NSM emphasizes the contributions of social determinants of health, such as rurality, social and community connections, and developmental needs.

There is significant evidence supporting the associations between SLEs and adverse maternal health outcomes, including PD. Relationship conflict and financial strain increase psychological distress, contributing to increased PD risk. Unintentional pregnancy may create conflict with significant others and is associated with increased stress and PD risk, particularly during the first months of pregnancy. ⁵ Furthermore, moving residential location is associated with increased PD risk, more so among those with more than one child. ⁶ Experiences of discrimination related to race and gender also significantly increase psychosocial distress, leading to increased PD risk. ⁷ The concurrent influence of SLEs and obstetrical risk (e.g., preterm labor, gestational diabetes, and hypertension) in rural settings further contributes to PD risk. ⁸ In contrast, social support can mitigate stress and is associated with decreased risk for PD.^6,8,9

Rurality, family household income, and educational status can disrupt the emotional and psychological balance of young families during a vulnerable time and are associated with increased occurrence of PD. ¹⁰ However, less is known about the relationship between SLEs and PD risk in rural populations in which economic circumstances, employment, and social connections are known to have a greater impact on perceived health risks compared to their urban and suburban counterparts. ² The NSM provides a multifactorial perspective to explain associations between SLEs and PD risk in rural populations. We sought to evaluate and describe the relationship between SLEs and PD risk in a rural population of diverse pregnant and postpartum individuals. We hypothesized that a higher number of SLEs would be associated with increased risk for PD, as measured by the Edinburgh Postnatal Depression Scale (EPDS).

Design/methods

As part of a larger parent study, we conducted a secondary cross-sectional analysis that included six rural Utah Health Districts and associated public health clinics. These clinics provide free public health services (e.g., vaccinations, nutrition services, and well-baby assessments) primarily to rural-dwelling low-income individuals, of whom 60%–70% are Medicaid recipients. All six rural health districts are designated as Health Professional Shortage Areas. A universal PD screening program was implemented in these health districts, providing study data from 2017 through 2024. Screening was offered to all pregnant and postpartum individuals visiting any one of the six rural public health clinics. The screening survey, administered via an electronic platform and self-completed by individuals, included basic demographics, pregnancy or postpartum status, preferred language (Spanish/English), the 10-item EPDS ¹¹ and an 8-item SLEs survey, ¹² and a 5-item depression and anxiety history survey. Funding was provided by the Utah Department of Health and two NIH larger trial grants (1R01NR017620-01A1 and 1RF1NR020841-01).

The 8-item SLE survey included questions about the immediate past 12 months, including job loss, partner’s job loss, inability to pay bills, partner’s pregnancy objection, conflicts with partner, separation or divorce, physical or sexual abuse, and moving place of residence. SLEs were examined individually and in groupings: 0 SLEs, 1 SLE, or ⩾2 SLEs in association with EPDS scores, consistent with analyses in other studies.^3,13,14 Individual EPDS scores equaled the sum of the 10 items (range 0–30). PD risk was defined as an EPDS score ⩾9. ¹⁵ All incomplete surveys (those with <70% completion) were removed from analysis. A proration approach was used to assign item values on missing EPDS data (2.2%). STROBE guidelines were used to guide this report. ¹⁶

Data analysis

Descriptive statistics, including frequency/percent and mean/standard deviation, are provided for demographic, SLEs, and EPDS data. We used multinomial logistic regression to evaluate the association between SLEs and PD risk and calculated odds ratios (OR) and 95% confidence interval (CI). We adjusted for a priori-selected variables; race (American Indian Alaskan Native, Asian, Black or African American, White, Native Hawaiian Pacific Islander, Multiple races, Other), Hispanic ethnicity (yes/no), age (years), pregnancy or postpartum status, partnered status (yes/no), type of insurance (Medicaid, none, private, and other), and whether screening took place before or after the COVID-19 pandemic (WHO declared a date of March 11, 2020).

Results

Demographics, SLEs, and EPDS

The data from 4245 completed screening surveys were included in the analysis. The most common racial backgrounds reported by individuals were white (79.8%, n = 3389), American Indian/Alaska Native (7.6%, n = 322), and “Other” (6.7%, n = 283). A total of 20.2% (856) of individuals self-identified as Hispanic.

The mean age of individuals was 26.8 years (standard deviation = 5.7), primarily married or living with a partner (n = 2772, 65.3%), and Medicaid recipients (2556, 60.2%) or with private insurance (925, 21.8%). Most individuals were postpartum (n = 2721, 64.1%) with 35.9% (n = 1524) being pregnant. A large proportion of individuals (49.4%) reported having one SLE and 10.5% reported two or more SLEs. Table 1 shows the frequency of reported SLEs occurring in the 12 months prior to screening. Of note, 24.1% of individuals reported having trouble paying bills, 14.7% argued with their partner more than usual, 12.5% reported job loss, and 25.9% moved in the past 12 months (Table 1). The mean EPDS score was 7.9 (6.1), and overall, 26.3% of individuals had a positive screen (⩾9 on EPDS).

Table 1.

Descriptive statistics for demographic characteristics, EPDS scores, and type and number of SLEs.

Demographic variable
Age as continuous	Freq (n = 4245)	%
Mean (SD)	26.8 (5.7)
Median [min, max]	26 [18, 51]
Missing	0
Age
Less than 24	1695	39.9
25–34	2089	49.2
35–44	450	10.6
45 or older	11	0.3
Pregnancy status
Pregnant	1524	35.9
Postpartum	2721	64.1
Race
American Indian/Alaska Native	322	7.6
Asian	38	0.9
Black or African American	45	1.1
NHPI	55	1.3
White	3389	79.8
Multiple races	113	2.7
Other	283	6.7
Ethnicity
Not Hispanic or Latino	3389	79.8
Hispanic or Latino	856	20.2
Marital
Single	1333	31.4
Married/living with partner	2772	65.3
Divorced	95	2.2
Widowed	17	0.4
Missing	28	0.7
Insurance type
None	491	11.6
Medicaid	2556	60.2
Private/group	925	21.8
Other	244	5.7
Missing	29	0.7
EPDS
Mean (SD)	7.9 (6.1)
Median [min, max]	7 [0, 30]
⩾9	1117	26.3
SLE in last 12 months
SLE moved	1100	25.9
SLE trouble paying bills	1021	24.1
SLE argued with partner more than usual	622	14.7
SLE job loss	530	12.5
SLE partner’s job loss	394	9.3
SLE divorce or separated	292	6.9
SLE history of physical or sexual abuse	259	6.1
SLE partner does not want you pregnant	107	2.5
Number of SLEs
0	1702	40.1
1	2099	49.4
2+	444	10.5

SLE: significant life event; EPDS: Edinburgh Postnatal Depression Scale; SD: standard deviation.

Logistic models using EPDS total score of 9 as binary cutoff

Using a clinically meaningful cutoff of 9 on the EPDS as a positive screen, multinomial logistic regression was utilized to examine the association of the number of SLEs (0, 1, 2+). Unadjusted odds ratios for the association between SLEs and PD were one SLE, OR = 1.93, 95% CI [1.66, 2.24], and two or more SLEs, OR = 7.03, 95% CI [5.91, 8.35].

When controlling for covariates including history of depression, age, race, ethnicity, partnered status, during COVID-19 or before, presence of insurance, and pregnancy status, adjusted odds were for the presence of one SLE (1.84, 95% CI [1.56, 2.15]) and two or more SLEs (5.18, 95% CI [4.31, 6.23]) compared to no SLEs (see Table 2). Only a history of depression was a significant covariate with a trend for positive screens during the COVID-19 pandemic (screens conducted on or after March 12, 2020). Stratified analysis by pregnancy status and COVID-19 revealed no significant changes in main predictors or covariates. As such, we report the parsimonious non-stratified analysis.

Table 2.

Logistic regression examining predictors for a positive EPDS screen (⩾9).

Predictor	B	SE	Sig.	Exp (B)	95% CI for Exp (B)
					Lower	Upper
0 SLEs (Reference)
1 SLE	0.607	0.082	<.001	1.835	1.563	2.153
2+ SLE	1.645	0.094	<.001	5.182	4.311	6.231
History of depression (no Hx reference)
History of depression	1.444	0.081	<.001	4.238	3.614	4.969
Race (White reference)
American Indian Alaskan native	0.017	0.135	0.902	1.017	0.78	1.325
Asian	−0.199	0.395	0.615	0.819	0.378	1.779
Black or African American	0.193	0.328	0.558	1.212	0.637	2.308
NHPI	−0.059	0.316	0.851	0.943	0.508	1.749
Multiple races	0.039	0.219	0.858	1.04	0.677	1.597
Other	−0.052	0.169	0.757	0.949	0.681	1.322
Ethnicity (non-Hispanic reference)
Hispanic	−0.147	0.106	0.164	0.863	0.701	1.062
Age (years)	−0.008	0.006	0.209	0.992	0.98	1.004
Pregnancy status (postpartum reference)
Pregnant	0.082	0.073	0.264	1.085	0.94	1.253
Relationship status (unpartnered reference)
Partnered	0.041	0.077	0.595	1.042	0.896	1.212
COVID-19 (pre-pandemic reference)
Post pandemic	−0.157	0.082	0.056	0.855	0.728	1.004
Insurance (Medicaid reference)			0.144
None	0.02	0.118	0.867	1.02	0.809	1.286
Private	0.046	0.091	0.615	1.047	0.876	1.252
Other	0.332	0.143	0.021	1.393	1.052	1.844
Intercept	−1.145	0.192	<.001	0.318

SLE: significant life event; EPDS: Edinburgh Postnatal Depression Scale; CI: confidence interval; SE: standard error.

Discussion

In this large sample of diverse rural-dwelling individuals, more than 26% had a positive PD screen (⩾9). The relationship between PD and SLEs was significant; individuals with one SLE had a nearly two-fold risk for PD, while those with two or more SLEs had a fivefold increase in risk. The most common SLEs reported were moving and trouble paying bills. This is similar to the findings in urban childbearing populations.^2,17,19

A strength of this study was the diverse population represented (e.g., Hispanic and Native American). Our analysis did not find that race/ethnicity contributed to PD risk. These findings support universal screening for SLEs in parallel with PD screening for all pregnant and postpartum individuals, and the provision of supportive resources in rural communities. As social support has a protective effect on PD risk,^8,9,18 encouraging and strengthening social networks in rural childbearing populations may help mitigate PD symptoms. ¹⁹

While the results are congruent with other studies,^2,3,17 a limitation of the cross-sectional study design is the inability to determine causality. In addition, self-reports are at risk for recall bias, particularly due to the 12-month length of recall for SLEs. In this study, family income, pregnancy complications, infant gender, and delivery mode were not known and, therefore, could not be accounted for in the analysis. This study’s results may not be generalizable to other rural populations. Future research should focus on longitudinal studies in other rural populations.

Approximately half of the sampling timeframe occurred after the COVID-19 pandemic began, which included a period of “lock down” in Utah. This may have impacted the occurrence of SLEs, particularly when viewed from the NSM. Although our study results did not identify a significant impact on SLEs during the pandemic, other studies reported an increase in stressors such as social isolation, financial instability, childcare demands, and reduced access to providers. These stressors can disrupt an individual’s flexibility and healthy functioning, thus contributing to increased risk for PD. ³ Furthermore, the switch to telehealth platforms and shelter-in-place orders during the pandemic may have resulted in increased anxiety and uncertainty. Results of data collected during the COVID-19 pandemic should be considered with caution, as gaps in screening make interpretation of results difficult.

Conclusion

This study’s findings reflect a diverse rural population and highlight the need for universal screening that includes SLEs and a history of depression, in addition to PD screening. Since environmental forces affect system stability and result in increased anxiety and uncertainty, rural persons with limited resources should be routinely assessed for baseline health and wellness and the condition of buffers that protect against stressors. Understanding the psychological impact of SLEs and the relationship with PD risk can contribute to improved perinatal care and health outcomes for all childbearing individuals living in rural communities, regardless of racial/ethnic background.