Navigating the Minority Grief Paradox: Racial and Ethnic Differences in Parental Death, Social Ties, and Early-Midlife Psychological Distress in the US

Abstract

The death of a parent can profoundly impact psychological and social functioning, particularly among racial-ethnic minority groups. Using data from the National Longitudinal Study of Adolescent to Adult Health (N = 9,508), the current study compared the association between parental death timing during different life stages and early-midlife depressive and perceived stress symptoms (“psychological distress”) across racial-ethnic identities, while accounting for early-midlife social support and strain. We also investigated the mediating effects of cumulative social support and strain between life course timing of first parental death and midlife psychological distress using the Karlson-Holm-Breen (KHB) method. We found that Black individuals experienced the highest frequency of parental death beginning earlier in life. Stepwise linear regression models showed that parental death during young adulthood and early midlife is associated with early-midlife psychological distress among White individuals, net of covariates. Our results among Black and Hispanic adults indicate that the life course timing of parental death was unrelated to early-midlife psychological distress, exposing an understudied area of the “Minority Mental Health Paradox”. Early-midlife social support and strain were strongly associated with early-midlife psychological distress across racial-ethnic identities. The KHB mediation analysis further revealed that cumulative social support in midlife independently accounted for approximately 16% and 20% of the association between parental death in young adulthood and midlife stress and depressive symptoms, respectively. Our study highlights that bereavement programs could tailor interventions to address variations in midlife social support and strain across different racial-ethnic groups to mitigate psychological distress.

Introduction

The death of a biological parent or parent figure is a stressful life event (hereafter referred to as “parental death”) that can undermine psychological wellbeing at any point in the life course (Stroebe & Stroebe, 2007). In 2021, 44.2% of the United States (US) population reported experiencing the death of at least one parent (Hayward, 2023). Approximately 4.3% of children aged 0-17 experienced parental death, compared to 11.5% and 23.4% of adults aged 18-29 and 30-39, respectively (Hayward, 2023). Experiencing parental death earlier in life has been linked to multiple markers of chronic stress, contributing to psychological challenges later in the life course (Luecken & Roubinov, 2012), including but not limited to, depressive symptoms and perceived stress (subsequently referred to as “psychological distress”) (Zhu et al., 2022).

Even so, early parental deaths during childhood and adolescence are disproportionately concentrated in lower income, minoritized communities in the US (Umberson et al., 2017). Non-Hispanic Black and Hispanic youth are more likely to experience parental death at a younger age compared to their non-Hispanic White counterparts (Umberson et al., 2017; Donnelly et al., 2023). Despite the heightened risk of parental death among systemically minoritized groups, limited research has examined whether the life course timing of parental death is differentially associated with early-midlife psychological distress across racial-ethnic groups (Donnelly, Lin, & Umberson, 2023). Therefore, our primary aim is to explore racial-ethnic differences between non-Hispanic White (“White”), non-Hispanic Black (“Black”), and Hispanic adults, in the association between the life course timing of parental death and early-midlife psychological distress.

Compounding potential psychological challenges, the grief response following parental death can alter an individual’s abilities to develop and maintain social relationships (Umberson & Donnelly, 2023). Social relationships with immediate family members – spouse or partner and child(ren) – are often significant sources of support or strain that may mitigate or exacerbate the negative health impacts of stressful life circumstances (Kim et al., 2021; Turner & Turner, 2013). As minoritized individuals are unduly exposed to parental death, they may also endure elevated social stress from the rippling effects of bereavement within their social network (Kim et al., 2021). Thus, the quality of surviving social relationships may disproportionately influence early-midlife psychological distress in the aftermath of parental death among systematically minoritized groups. Indeed, social support and strain in one’s surviving relationships may serve as mediating pathways connecting parental death and early-midlife psychological distress (Lincoln et al., 2005). Therefore, our secondary aim is to examine racial-ethnic variation in the roles of social support and strain on the association between life course timing of parental death and early-midlife psychological distress.

By examining race-ethnic variations in the association between the life course timing of parental death and early-midlife psychological distress, and the role of social relationship quality, we advance the literature in three ways. First, we clarify racial-ethnic variations in the salience of biological parent and parental figure death as a source of psychosocial disadvantage. Further, whereas existing studies have focused on Black-White differences in the health implications of parental death (Garcia et al., 2024), our study also examines Hispanic adults, a rapidly increasing minority population in the US (Barrett, 2020). Second, we identify the extent to which social relationships influence the magnitude of early-midlife psychological distress following parental death for different racial-ethnic groups. Third, we assess whether the accumulation of social support and strain are mediating pathways between parental death and early-midlife psychological distress.

Background

Conceptualizing racial-ethnic differences in parent death and psychological distress

The life course and stress process perspectives (Elder Jr., 1998; Pearlin, 2010) frame our examination of racial-ethnic variation in the life course timing of parental death and early-midlife psychological distress, and the potential role of social support and strain on this association. Specifically, we draw on three key principles shared by both theoretical perspectives: 1) social stratification that differentially influences individual experiences of stressors and resources, 2) timing of life events, and 3) linked lives across the life course (Elder Jr, 1998; Pearlin, 2010).

We posit that racial-ethnic differences in the association between the life course timing of parental death and early-midlife psychological distress reflect broader systemic inequities that disadvantage minoritized groups in the US (Williams & Sternthal, 2010; Douglas et al., 2021). Black and Hispanic individuals have a higher risk than White individuals of experiencing family member death earlier in life (Umberson, 2017). Racial-ethnic minorities face a heavier toll of structural disadvantages and stressors over their life course, which can reduce internal and external resources necessary to cope following parental death (Umberson, 2017). Black and Hispanic children and adults also have lower access to and utilization of mental health care services, even during periods of crisis (Marrast, Himmelstein & Woolhandler, 2016). Therefore, the confluence of structural challenges and inadequate healthcare results in limited resources for minoritized communities to manage bereavement-related stress (Kim et al., 2021).

Despite higher overall stress exposure, Black and Hispanic adults have a lower prevalence of mental health disorders and psychological distress compared to White adults, commonly known as the “Minority Mental Health Paradox” (Kiecolt et al., 2008; Hernandez et al., 2022). Some scholars have suggested that racial-ethnic minority groups may develop different stress appraisal and cognitive strategies to adapt to and protect against proliferating stressors throughout the life course (Ng-Mak et al., 2002; Brown, Mitchell & Ailshire, 2020). For instance, coping mechanisms such as desensitization, may explain how minoritized communities exposed to chronic stressors may adapt to their circumstances by suppressing emotional distress during difficult life events (Ng-Mak et al., 2002). Other research suggests that spirituality or maintained bonds with deceased family members may also buffer against negative grief outcomes among Black individuals (Laurie & Neimeyer, 2008). Thus, minoritized individuals may have developed coping strategies from an early age to manage greater stress exposure, which may contribute to greater resilience in the face of stressful events, like parental death (Williams et al., 2010).

These paradoxical effects have been documented across various mental health conditions, including mood, stress, and substance use disorders, and have been consistently observed across different age groups and genders (Erving, Thomas, & Frazier, 2019). To date, few studies have examined this phenomenon within the context of bereavement (Denckla et al., 2022). One study found that the risk for any lifetime psychiatric disorder was not significantly moderated by race or ethnicity among parentally bereaved youth (Denckla et al., 2022). Yet, the combination of structural (dis)advantages throughout the life course, and differences in coping strategies, suggests that the experience of parental death and links to psychological distress is likely quantitatively and qualitatively distinct across Whites, Blacks, and Hispanic adults.

Life course timing of parental death and mental health

Life course and stress process perspectives suggest that the impact of specific life events on health and other social processes will vary depending on the age at which they occur (Elder Jr, 1998; George, 2013). Compared to later life stages, negative early life experiences often disrupt psychological and emotional regulation, thereby setting the stage for worse psychological wellbeing over time (Avison, 2010). Research has repeatedly shown that parental death during childhood and adolescence (ages 0 to 9, and 10 to 20, respectively) can contribute to worsened physical and mental health due to developmental susceptibility and material dependency unique to this age group (Stroebe & Stroebe, 2007; McKay et al., 2021). Experiencing the death of a parent earlier in life leads to an extended timeframe for other risk factors to develop, ranging from behavioural, economic, and social (Donnelly, Lin, & Umberson, 2023). For instance, bereaved youth have a heightened risk of rapid uptake of adult roles (e.g., precocious transitions), which may further reinforce lifelong psychological challenges (Farella Guzzo & Gobbi, 2023). Overall, the death of a parent can significantly redirect multiple aspects of a youth's life, with the resulting instability and hardships playing a crucial role in shaping their later psychological outcomes (Donnelly, Lin, & Umberson, 2023).

Conversely, some scholars have proposed that parental death in young adulthood and early-midlife may still pose coping challenges due to continued parental dependence (Kamis, Stolte, & Copeland 2012; Donnelly et al., 2023). For instance, parents typically offer crucial support to young adults (ages 18 to 27) through financial assistance or emotional reassurance, all of which are essential for personal and professional development (Farella, Guzzo, & Gobbi, 2023). Similarly, in early-midlife (ages 30 to 45), balancing the demands of paid work and family obligations may also necessitate social support from parents (Infurna et al., 2020). Indeed, one study found that biological father death during young adulthood was uniquely associated with higher depressive symptoms in middle adulthood, suggesting that the effects of parental death likely depend on life course timing (Kamis, Stolte, & Copeland, 2022).

Linked Lives in the Context of Parental Death

The concept of linked lives extends intergenerationally with the transmission of biological and social antecedents from parents influencing the wellbeing and development of offspring (Elder Jr., 1998). Therefore, parental death is a significant stressor that can impair health directly from the permanent loss of a close social tie and intergenerational bond (Bengtson, Elder Jr. & Putney, 2012), and indirectly through stress proliferation (Pearlin, 2010). Secondary stressors may include social isolation (Umberson & Donnelly, 2023) and strained relationships with surviving family members, including siblings (Khodyakov & Carr, 2009), children (Kim et al., 2021), and spouses (Stokes, 2016). Poor quality social relationships may reduce one’s own capacity to manage bereavement stress and increase risks of depressive symptoms and stress beyond typical trajectories of grief (Umberson & Donnelly, 2023). Nonetheless, parental death can also activate support from kin and non-kin – even alongside social strain – that can attenuate the psychological and social challenges of bereavement (Thomas, 2016; Shin & Gyeong, 2023; Stokes, 2016).

Critically, the associations between social support and strain and psychological distress also vary across racial-ethnic groups (Lincoln et al., 2003). Among minoritized groups, receiving social support from family members and other informal social networks is more strongly associated with lower depressive symptoms (Lincoln et al., 2005). Negative social interactions with family members have been associated with immediate psychological distress for both White and Black adults but are associated with a higher risk of lifetime anxiety and depression among older Black adults only (Lincoln et al., 2010). Within the context of stressful life events, one study found that middle-aged Black adults tend to experience more relationship strain, specifically with adult children, relative to White adults following parental death (Kim et al., 2021). Another study by Lincoln and colleagues (2003) examined the impact of traumatic life events (excluding parental death) on psychological distress, and the mediating roles of support and negative interactions, across White and Black adults. Among White adults, traumatic events were linked to increased psychological distress, negative social interactions with family, and lower social support, with the latter two mediating this relationship. In contrast, Black adults experienced direct associations between traumatic events and negative social interactions and psychological distress, but there was no mediation effect from social support. While this seminal study suggests that traumatic life events may be more impactful on family relationship quality and psychological distress among White than Black adults, more research is needed to understand how parental death differentially shapes these interconnected relationships across racial-ethnic groups.

The Current Study

Building on the existing evidence base, the current study investigates the following research questions:

Research Question 1:

Are there racial-ethnic differences in the association between life course timing of parental death and early-midlife psychological distress? Compared to non-bereaved adults, we hypothesize that early life course experiences of parental death will be associated with increased psychological distress (Hypothesis 1) and this association will be less pronounced for Black and Hispanic adults relative to White adults (Hypothesis 2).

Research Question 2:

a) Do early-midlife social support and strain attenuate or amplify the magnitude of the association between the life course timing of parental death and mental health? b) If so, are there racial-ethnic differences in the role of social support and strain on early-midlife psychological distress? We hypothesize that cumulative social support and strain will respectively attenuate and strengthen the association between life course timing of parental death and psychological distress (Hypothesis 3), and this effect will be stronger for Black and Hispanic adults relative to White adults (Hypothesis 4).

Research Question 3:

a) Does perceived social support and strain in early-midlife serve as a mediator between life course timing of first parent death experience and early-midlife psychological distress? We hypothesize that cumulative social support and strain will both independently mediate the association between life course timing of parental death and psychological distress (Hypothesis 5).

Data and Methods

Data and sample selection

The current study used restricted data from the National Longitudinal Study of Adolescent to Adult Health, otherwise known as “Add Health” (Harris, 2013). Add Health is a nationally representative sample of 20,745 adolescents in grades 7–12 in 1994–1995 in the US that used a multi-stage, school-based cluster design by selecting 80 high schools and a paired feeder school from a list of all high schools in the US in 1994-1995. This cohort has been followed for six waves to date beginning with Wave I (1994-1995; grades 7-12), Wave II (1995–1996; grades 8-12), Wave III (2001; ages 18–26 years), Wave IV (2008–2009; ages 24–32 years), Wave V (2016–2018; ages 33–43 years) and Wave VI (2022-2025; ages 39-51 years). Additional details about the Add Health study are described elsewhere (Harris, 2013).

The eligible sample for the current study consisted of US adults who had cross-sectional Wave V weights (N=12,300). Participants were excluded if they did not provide valid data on biological parent or parent figure death (N = 746), depressive symptoms or perceived stress (N = 295), availability of social ties (n = 209), relationship support and strain measures (n = 189), as well as study covariates (n = 1,353). The final analytic sample (N = 9,508) was stratified into racial ethnic subsamples: White (N = 6,221), Black (N = 1,840), and Hispanic (N = 1,447).

Measures

Dependent Variables

Depressive Symptoms:

The first outcome was measured using the 5-item “Negative Affect” subscale of the Centre for Epidemiological Studies Depression Scale (CES-D) in Wave V (Carleton et al., 2013). The intended use of the 5-item CES-D subscale is to understand the prevalence of state-specific depressive symptoms within a population (Bohannon et al., 2003). Respondents shared how often in the past week they experienced depressive symptoms: (1) could not shake off the blues, even with help from family and friends; (2) felt depressed during the past 7 days, (3) felt sad, (4) happy (reverse coded), and (5) felt life was not worth living. The response options for each item included 0 (Never, 0 days), 1 (Rarely, 1-2 days), 2 (Occasionally, 3-4 days), and 3 (severe, 5-7 days). All items were summed to create a continuous score ranging from 0-15, with higher scores indicating experiencing worse depressive symptoms. The Cronbach’s alpha of 0.88 demonstrated good reliability in the current sample.

Perceived Stress:

The second outcome was another state-specific measure of coping using the four-item Perceived Stress Scale (PSS-4) included in Wave V (Cohen, Kamarck & Mermelstein, 1983). The PSS-4 evaluates the degree to which individuals believe their life has been unpredictable or uncontrollable over the past 30 days. Participants rated, on a Likert-type scale ranging from “never” to “very often”, the frequency with which they 1) “felt unable to control their lives,” 2) “confident in their ability to handle problems,” 3) “that things were going their way,” and 4) “unable to overcome difficulties”. All items were summed to create a cumulative score ranging from 0–16, with higher scores demonstrating worse perceived stress (Cohen, Kamarck & Mermelstein, 1983; Lee, 2012). The Cronbach’s alpha indicated acceptable reliability in the current sample (α = 0.79).

Independent variables

Life Course Stage of Parental Death(s):

The primary independent variable includes the death(s) of biological mothers and fathers as well as mother and father figures assessed in Wave V. All respondents answered the question “We would like to know about the woman [man] you feel raised you. Would you say your mother [father] figure was your biological mother [father] or some other woman [man]?” with response options 1) biological mother [father], 2) other mother [father] figure, and 3) no mother [father] figure. Parental death was ascertained with the subsequent items “Is your biological [mother or father] or mother [father] figure still alive?”] with a binary (0) No and (1) Yes response. Additionally, participants provided the month and year of the biological parent or parent figure's death, if applicable. Guided by prior studies (Garcia et al., 2024), respondents’ age at the parental death was recoded into three life stage categories: childhood (0-17 years), young adulthood (18-29 years), and early-midlife (30-43 years), compared to individuals who did not experience parental death. The original conceptualization of childhood (0-9 years) and adolescence (10-17 years) lacked sufficient sample size to serve as standalone life stage categories. For those with multiple deaths of biological parents and/or parent figures across multiple life course stages, individuals were categorized into the life stage of their first experience of a parental death. Appendix Table A1 provides additional description of variable operationalization.

Race & Ethnicity:

The constructed race variable available in Wave I was used to assign participants into one single race category (White, Black, Native American, or Asian) (Udry, Li & Hendrickson-Smith, 2003). To account for ethnicity, respondents were also asked “Are you of Hispanic or Latino origin?” with a binary response in Wave I. We derived a new variable by combining racial-ethnicity into the following mutually exclusive categories of (1) non-Hispanic, White, (2) non-Hispanic, Black, and (3) Hispanic. All other racial categories were dropped due to statistical power concerns.

Mediators

Perceived Social Support:

In Wave V, participants reported their perceived support from their social network including, 1) spouse/partner, 2) child(ren), 3) extended family, and 4) friends based on two questions: “please indicate whether or not you can open up to [them] if you need to talk about your worries”, and “please indicate whether or not you can rely on [them] for help if you have a problem” with each item having binary response option (0) No and (1) Yes. We constructed a cumulative social support score that ranged between 0-8, with higher scores indicating higher social support across more relationships.

Perceived Social Strain:

In Wave V, participants reported their perceived level of strain with their 1) spouse/partner, 2) child(ren), 3) extended family members, and 4) friends. Participants were asked, “please indicate whether or not [they] ever make too many demands or criticize you” with the binary response option (0) No and (1) Yes. We constructed a cumulative social strain variable that ranged from 0-4, with higher scores indicating higher perceived social strain across more relationships.

Covariates

Analyses adjusted for participant- and family-level social covariates collected at various data collection waves Participant-level covariates included current age (years), sex (0 = female, 1 = male), racial-ethnic(0 = White, 1 = Black, 2 = Hispanic) (in full sample analyses only), general self-rated health at age 16 (0 = Excellent/Very Good/Good, 1 = Fair/Poor), and nativity status (0 = US born, 1 = foreign born) collected in Waves I-V. Family-level covariates included parent disability status (0 = not disabled, 1 = disabled) as well as the “Social Origin Score”, which was constructed using parental education, parental occupation, household income, and household receipt of public assistance from the Wave I Parent Survey (Belsky, Domingue and Harris, 2020). We also controlled for the availability of social ties including 1) presence of spouse or partner (0 = not married, cohabitating, or dating a romantic partner, 1 = married, cohabitating, or dating a romantic partner), 2) presence of child(ren) (0 = no child(ren) living, 1 = at least one current child living), 3) presence of extended family (0 = no extended family, 1 = presence of extended family), and 4) presence of close friends (0 = no close friends, 1 = one or more close friends).

Statistical Analysis

Statistical analyses were performed using Stata 18.0 SE. First, we conducted descriptive statistics on all study variables stratified by racial-ethnic group (Table 1). Figure 1 presents a visual framework of the analytic plan that compares racial-ethnic differences in life course timing of parental death, social ties, and early-midlife psychological distress described herein.

Table 1.

Weighted descriptive statistics of final analytic sample stratified by race and ethnicity (N = 9,508)

	Full Sample	White	Black	Hispanic	p-value
	Mean or %	Mean or %	Mean or %	Mean or %
Current Age (years), Mean (SE)	37.79 (0.12)	37.72 (0.13)	38.04 (0.21)	37.91 (0.23)	0.33
Biological sex					0.93
Female	50.36%	50.28%	50.16%	51.05%
Male	49.64%	49.72%	49.84%	48.95%
Nativity status					<0.001 ***
Born outside the US	4.16%	1.20%	3.05%	23.43%
Born in the US	95.84%	98.80%	96.95%	76.57%
Social origin score, M (SE)	0.09 (0.06)	0.32 (0.06)	−0.47 (0.11)	−0.62 (0.10)	<0.001 ***
Perceived childhood general health					0.42
Excellent/Very Good/Good	95.81%	96.05%	95.49%	94.70%
Fair/Poor	4.19%	3.95%	4.51%	5.30%
Parent disability during childhood					0.41
Not disabled	92.35%	92.69%	91.36%	91.49%
Disabled	7.65%	7.31%	8.64%	8.51%
Parent death ever					<0.001 ***
No death	70.50%	72.84%	57.07%	72.28%
Yes	29.50%	27.16%	42.93%	27.72%
Deceased parent identity					<0.001 ***
Biological Mother	9.22%	8.05%	16.36%	7.85%
Biological Father	18.86%	17.57%	25.31%	18.96%
Mother Figure	2.49%	2.02%	5.09%	2.25%
Father Figure	3.98%	3.50%	7.09%	3.14%
Number of parent deaths ever					<0.001 ***
No death	70.50%	72.84%	57.07%	72.28%
One death	24.81%	23.37%	33.15%	23.61%
Two or more deaths	4.69%	3.79%	9.78%	4.12%
Gender of deceased parent(s)					<0.001 ***
No death	70.50%	72.84%	57.07%	72.28%
Only maternal death(s)	7.42%	6.68%	12.17%	6.24%
Only paternal death(s)	18.15%	17.30%	22.62%	18.03%
Both maternal and parental death(s)	3.93%	3.18%	8.14%	3.46%
Life stage of first parent death					<0.001 ***
No death	70.50%	72.84%	57.07%	72.28%
Childhood (0-17 years)	5.18%	4.05%	9.06%	7.47%
Young Adulthood (18-29 years)	10.64%	10.05%	15.96%	7.91%
Early Midlife (30-43 years)	13.68%	13.06%	17.77%	12.34%
Number of life stages with parent death(s)					<0.001 ***
No death	70.50%	72.84%	57.07%	72.28%
One life stage	26.67%	24.75%	37.21%	25.80%
Two or more life stages	2.83%	2.41%	5.72%	1.93%
Involvement of deceased parent(s) during childhood					<0.001 ***
No death	70.50%	72.84%	57.07%	72.29%
Raised participant	23.68%	22.48%	32.35%	20.66%
Did not raise participant	4.57%	3.74%	7.69%	5.89%
Combination raised and did not raise participant	1.25%	0.94%	2.90%	1.16%
Availability of social ties
Married or partnered	84.70%	85.98%	78.65%	84.14%	<0.001 ***
Has living child(ren)	69.67%	70.07%	68.15%	69.05%	0.54
Connected to extended family	95.72%	96.92%	91.49%	93.48%	<0.001 ***
Has close friends	94.63%	95.45%	91.35%	93.53%	<0.001 ***
Quality of Social Ties
Overall support (Range 0-8), M (SE)	5.76 (0.04)	5.90 (0.03)	5.29 (0.09)	5.51 (0.09)	<0.001 ***
Spouse support					<0.001 ***
0 (None/Low)	20.13%	17.82%	29.81%	22.66%
1 (Moderate)	5.07%	4.47%	7.84%	5.42%
2 (High)	74.79%	77.71%	62.35%	71.92%
Child support					0.12
0 (None/Low)	46.97%	47.69%	42.89%	47.44%
1 (Moderate)	15.13%	14.34%	17.52%	17.11%
2 (High)	37.90%	37.97%	39.60%	35.45%
Extended family support					<0.001 ***
0 (None/Low)	10.21%	8.58%	16.22%	12.97%
1 (Moderate)	14.68%	13.17%	19.61%	17.98%
2 (High)	75.11%	78.26%	64.17%	69.05%
Friend Support					<0.001 ***
0 (None/Low)	10.93%	9.34%	16.17%	14.34%
1 (Moderate)	12.22%	11.17%	16.05%	14.05%
2 (High)	76.85%	79.49%	67.78%	71.61%
Overall strain (Range 0-4), M (SE)	0.73 (0.01)	0.72 (0.02)	0.82 (0.04)	0.71 (0.03)	0.04 *
Spouse strain					0.62
0 (None)	77.20%	77.57%	76.05%	76.33%
1 (Any)	22.80%	22.43%	23.67%	23.67%
Child strain					0.04 *
0 (None)	87.17%	86.70%	87.14%	90.26%
1 (Any)	12.83%	13.30%	12.86%	9.74%
Extended family strain					0.18
0 (None)	73.00%	73.46%	70.42%	73.30%
1 (Any)	27.00%	26.54%	29.58%	26.70%
Friend strain					<0.002 **
0 (None)	89.30%	90.13%	84.43%	89.11%
1 (Any)	10.70%	9.87%	15.47%	10.89%
Mental health outcomes
Depressive symptoms (0-15), M (SE)	2.87 (0.05)	2.83 (0.06)	3.10 (0.14)	2.83 (0.13)	0.21
Perceived stress (0-16), M (SE)	5.01 (0.05)	4.92 (0.06)	5.40 (0.12)	5.03 (0.11)	0.002 **

Note: * p<0.05, ** p≤0.01, *** p≤0.001

M (SE)= Mean, Standard Error

Figure 1. Analytic framework conceptualizing racial-ethnic variation in life course timing of parental death, social ties, and early-midlife psychological distress.

Note: Numbers in figure refer to each sequential analytic step in current study

To test the main study hypotheses in Step 1, we conducted separate stepwise multivariate linear regression models to assess the association between life stage of the first parental death to early-midlife psychological distress (e.g., depressive symptoms and perceived stress) in Wave V. Tables 2 and 3 present fully adjusted models incorporating social covariates, availability of social ties, and cumulative social support and strain for the full sample and each racial ethnic sub-sample. We also modelled the effects of each individual social tie to distinguish differential effects on stress and depressive symptoms in early-midlife. Supplementary Tables 2 - 9 provide stepwise model results (Woodward, Quashie, & Zhang, 2026). For Step 2, we examined the association between the life stage of first parental death and cumulative perceived support and strain scores in Wave V for preparation for mediation analysis. We conducted unadjusted and adjusted models to establish if there was variation in social support and strain by life course timing of first parental death and by racial-ethnic group (Supplementary Tables 10 - 13) (Woodward et al., 2026). Finally, in Step 3, we used the Karlson-Holm-Breen (KHB) method to examine cumulative support and strain as mediators in the relationship between life course stage of first parental death and early-midlife psychological distress (e.g., depressive symptoms and perceived stress) in the full analytic sample (Table 4) (Karlson, Holm, & Breen, 2012). We only examined first parental death during young adulthood, given that there were no statistically significant associations with other life course stages and cumulative social support and strain. The KHB method allows for simultaneous testing of multiple correlated mediators while decomposing the unique contribution of each included mediator (Karlson, Holm, & Breen, 2012). Notably, the KHB provides percentage change in the key independent variable after adding mediator(s) as well as statistical significance of each mediator with a z-score calculation. All regression models accounted for nonresponse and study design effects with Add Health Wave V cross-sectional sample weights and significance level was set at p = 0.05.

Table 2.

Linear Regression Estimates of Timing of Parental Death on Midlife Perceived Stress, Stratified by Race/Ethnicity (N = 9,508)

	Full Sample		White		Black		Hispanic
	N = 9,508		n = 6,115		n = 1,788		n = 1,414
Variables	b	95% CI	b	95% CI	b	95% CI	b	95% CI
Life stage of first parent death (Ref = no parental death)
Early Midlife	0.298*	(0.023 - 0.574)	0.353*	(0.028 - 0.678)	0.016	(−0.554 - 0.587)	0.317	(−0.516 - 1.151)
Young Adulthood	0.381*	(0.091 - 0.671)	0.408*	(0.047 - 0.770)	0.455	(−0.252 - 1.162)	0.139	(−0.498 - 0.777)
Childhood	−0.071	(−0.400 - 0.259)	0.093	(−0.382 - 0.569)	0.138	(−0.412 - 0.688)	−0.856	(−1.842 - 0.130)
Current Age (years)	−0.037	(−0.080 - 0.007)	−0.038	(−0.090 - 0.014)	−0.036	(−0.147 - 0.076)	−0.055	(−0.160 - 0.051)
Biological Sex (ref = Female)	−0.563***	(−0.730 - −0.395)	−0.621***	(−0.805 - −0.437)	0.016	(−0.554 - 0.587)	−0.401	(−0.818 - 0.015)
Race/Ethnicity (ref = White)	−0.073	(−0.192 - 0.046)	--	--	--	--	--	--
Nativity Status (ref = US born)	−0.432	(−0.990 - 0.127)	0.442	(−0.579 - 1.463)	−0.380	(−1.510 - 0.750)	−0.785*	(−1.468 - −0.102)
Social Origin Score	−0.186***	(−0.253 - −0.120)	−0.230***	(−0.308 - −0.152)	−0.037	(−0.180 - 0.107)	−0.104	(−0.286 - 0.078)
Self-rated Health at 16 (Wave V)	0.670**	(0.243 - 1.096)	0.803**	(0.313 - 1.293)	0.953	(−0.128 - 2.034)	0.031	(−0.978 - 1.039)
Parent disability in childhood	0.144	(−0.180 - 0.468)	−0.030	(−0.368 - 0.309)	0.069	(−0.703 - 0.842)	0.967	(−0.026 - 1.961)
Availability of Social Ties
Married or partnered	−0.281	(−0.582 - 0.021)	−0.284	(−0.658 - 0.090)	−0.493	(−1.039 - 0.053)	0.308	(−0.480 - 1.097)
Has living child(ren)	0.372**	(0.150 - 0.594)	0.317*	(0.041 - 0.593)	0.621*	(0.122 - 1.121)	−0.103	(−0.653 - 0.447)
Has extended family members	−0.148	(−0.596 - 0.299)	0.055	(−0.537 - 0.646)	−0.230	(−1.009 - 0.550)	−0.265	(−1.156 - 0.626)
Has close friends	−0.555**	(−0.960 - −0.149)	−0.750**	(−1.208 - −0.291)	−0.124	(−1.162 - 0.915)	0.289	(−0.886 - 1.463)
Cumulative Social Support	−0.431***	(−0.492 - −0.369)	−0.442***	(−0.518 - −0.366)	−0.374***	(−0.489 - −0.258)	−0.785*	(−1.468 - −0.102)
Cumulative Social Strain	0.722***	(0.648 - 0.796)	0.738***	(0.659 - 0.817)	0.544***	(0.357 - 0.731)	−0.104	(−0.286 - 0.078)
Model R-Squared	0.18		0.19		0.14		0.23

Note: *** p<0.001, ** p<0.01, * p<0.05

Table 3.

Linear Regression Estimates of Timing of Parental Death on Midlife Depressive Symptoms, Stratified by Race/Ethnicity Table (N = 9,508)

	Full Sample		White		Black		Hispanic
	N = 9,508		n = 6,115		n = 1,788		n = 1,414
Variables	b	95% CI	b	95% CI	b	95% CI	b	95% CI
Life stage of first parent death (Ref = no parental death)
Early Midlife	0.465***	(0.217 - 0.713)	0.451**	(0.161 - 0.741)	0.387	(−0.285 - 1.058)	0.664	(−0.002 - 1.330)
Young Adulthood	0.340*	(0.069 - 0.612)	0.377*	(0.060 - 0.695)	0.081	(−0.401 - 0.563)	0.537	(−0.399 - 1.473)
Childhood	0.082	(−0.241 - 0.405)	0.290	(−0.181 - 0.762)	−0.263	(−0.796 - 0.269)	−0.119	(−0.996 - 0.758)
Current Age (years)	−0.006	(−0.045 - 0.032)	0.005	(−0.038 - 0.049)	−0.063	(−0.172 - 0.046)	−0.007	(−0.114 - 0.099)
Biological Sex (ref = Female)	−0.408***	(−0.571 - −0.245)	−0.462***	(−0.643 - −0.282)	−0.305	(−0.712 - 0.101)	−0.257	(−0.711 - 0.197)
Race/Ethnicity (ref = White)	−0.131	(−0.270 - 0.008)	--	--	--	--	--	--
Nativity Status (ref = US born)	−0.246	(−0.708 - 0.217)	0.960	(−0.076 - 1.996)	−0.735	(−1.586 - 0.116)	−0.592*	(−1.118 - −0.067)
Social Origin Score	−0.097**	(−0.159 - −0.035)	−0.154***	(−0.234 - −0.075)	0.019	(−0.095 - 0.134)	−0.038	(−0.220 - 0.144)
Self-rated Health at 16 (Wave V)	1.011***	(0.420 - 1.601)	1.028**	(0.314 - 1.741)	2.093**	(0.706 - 3.480)	−0.191	(−0.956 - 0.574)
Parent disability in childhood	0.391*	(0.064 - 0.717)	0.273	(−0.096 - 0.643)	0.327	(−0.377 - 1.031)	0.996	(−0.275 - 2.267)
Availability of Social Ties
Married or partnered	−0.264	(−0.569 - 0.041)	−0.265	(−0.611 - 0.081)	−0.260	(−0.827 - 0.306)	−0.245	(−1.191 - 0.701)
Has living child(ren)	0.146	(−0.063 - 0.354)	0.142	(−0.111 - 0.395)	0.273	(−0.253 - 0.800)	0.116	(−0.395 - 0.628)
Has extended family members	−0.079	(−0.539 - 0.380)	−0.097	(−0.745 - 0.552)	−0.302	(−0.951 - 0.347)	−0.023	(−0.951 - 0.905)
Has close friends	−0.433*	(−0.853 - −0.012)	−0.476	(−0.979 - 0.027)	−0.558	(−1.237 - 0.121)	−0.049	(−1.276 - 1.178)
Cumulative Social Support	−0.431***	(−0.494 - −0.369)	−0.453***	(−0.525 - −0.382)	−0.402***	(−0.528 - −0.276)	−0.373***	(−0.524 - −0.221)
Cumulative Social Strain	0.614***	(0.526 - 0.702)	0.598***	(0.495 - 0.701)	0.594***	(0.388 - 0.800)	0.705***	(0.451 - 0.958)
Model R-Squared	0.17		0.18		0.18		0.17

Note: *** p<0.001, ** p<0.01, * p<0.05

Table 4.

KHB Test for Mediators Between Parental Death During Young Adulthood and Psychological Outcomes: Total Sample (N = 9,508)

	Perceived Stress		Depressive Symptoms
Mediators	% Reduction	z-Score	% Reduction	z-Score
Cumulative Social Support	15.96%	2.03*	19.57%	2.05*
Cumulative Social Strain	2.79%	0.30	2.93%	0.32
Total	18.75%	–	22.50%	–

Note: *** p<0.001, ** p<0.01, * p<0.05

All models controlled for social covariates (age, biological sex, nativity, social origin score, perceived childhood general health, parent disability) and availability of social ties (has spouse, child(ren), extended family, and friends).

Results

Descriptive characteristics

Table 1 shows that stratification by racial-ethnic group revealed statistically significant differences across key variables and social covariates. There were statistically significant differences in nativity status (i.e., 23.43% Hispanic participants born outside of the US) and average social origin score, with White adults having the highest score (M = 0.32, SE = 0.06) compared to Hispanic adults having the lowest (M = −0.62, SE = 0.10). We also observed significantly higher scores for perceived stress (M = 5.40, SD = 0.12) among Black adults compared to both White (M = 4.92, SD = 0.06) and Hispanic (M = 5.03, SD = 0.11) adults. While the average depressive symptom score for Black adults was higher (M = 3.10, SD = 0.14), it was not statistically significantly different from White (M = 2.83, SD = 0.06) and Hispanic (M = 2.83, SD = 0.13) adults. Trends also diverged with respect to the availability and quality of social ties in early-midlife, with Black adults experiencing the lowest cumulative score of social support (Black M = 5.29, SD = 0.09 versus White M = 5.90, SE = 0.03 versus Hispanic M = 5.51, SE = 0.091) and the highest cumulative score of social strain (Black M = 0.82, SD = 0.04 versus White M = 0.72, SE = 0.02 versus Hispanic M = 0.71, SE = 0.03) relative to other racial-ethnic groups. Overall, Black adults experienced the lowest levels of social support and highest levels of social strain across all individual relationships, except with their living children.

Regarding first parental death across racial-ethnic identities, White and Hispanic adults experienced a similar frequency of parental death by Wave V (27.17% versus 27.72%) but were eclipsed by Black adults (42.93%). Disparities in first experience of parent death were most pronounced during childhood (4.05% White versus 7.47% Hispanic versus 9.06% Black). Differences in in parent death remained consistent in later life course stages as Black adults reported a higher frequency in young adulthood (15.96%) and early-midlife (17.77%), relative to White and Hispanic adults. Furthermore, Black adults experienced more parental deaths (i.e., 9.78% exposed to two or more deaths) across multiple life stages (i.e., 5.72% two or more life stages) compared to other racial-ethnic groups. Therefore, parent death was experienced earlier, more frequently, and more consistently throughout the life course for Black adults.

Multivariate results

Early Early-midlife Perceived Stress Symptoms

Table 2 presents the fully adjusted multivariate linear regression results for the association between life course stage of parental death and perceived stress for the full sample and by racial-ethnic group. While the effect of parental death in childhood was null (b = −0.07, 95% CI: −0.40 – 0.26), parental death in young adulthood (b = 0.38, 95% CI: 0.09 – 0.67) and early-midlife (b = 0.30, 95% CI: 0.02 – 0.57) were positively and statistically significantly associated with early-midlife perceived stress compared to no parental death in the full sample. White adults who experienced their first parental death during young adulthood (b = 0.41, 95% CI: 0.05 - 0.77) and early-midlife (b = 0.35, 95% CI: 0.03 - 0.68) had significantly higher perceived stress in early-midlife compared to non-bereaved White adults. Parental death at any life course stage was not significantly associated with perceived stress in early-midlife among Black or Hispanic adults.

Cumulative social support was associated with lower stress symptoms (b = −0.43, 95% CI: −0.49 to −0.37), while cumulative strain was associated with higher stress symptoms (b = 0.72, 95% CI: 0.65 to 0.80) in the full sample. This trend was mirrored across all racial-ethnic subsamples: White adults (Cumulative Social Support b = −0.44, 95% CI: −0.52 - −0.37, Cumulative Social Strain: b = 0.74, 95% CI: 0.66 - 0.82), Black adults (Cumulative Social Support b = −0.37, 95% CI: −0.49 - −0.26, Cumulative Social Strain: b = 0.54, 95% CI: 0.36 - 0.73), and Hispanic adults (Cumulative Social Support b = −0.49, 95% CI: −0.58 - −0.32, Cumulative Social Strain: b = 0.85, 95% CI: 0.64 - 1.06). Overall, adjusting for cumulative sources of social support and strain attenuated the association between first parental death and perceived stress in early-midlife, except among Hispanic adults. adults.

Early-midlife Depressive Symptoms

The full sample showed a significant positive association between first parental death and depressive symptoms during young adulthood (b = 0.34, 95% CI: 0.07 to 0.61) and early-midlife (b = 0.47, 95% CI: 0.22 to 0.71), yet not during childhood (b = 0.08, 95% CI: −0.24 to 0.41) (Table 3). When looking at racial-ethnic subsamples, parental death during young adulthood (b = 0.38, 95% CI: 0.06 – 0.70) and early early-midlife (b = 0.45, 95% CI: 0.16 – 0.74) was significantly and positively associated with depressive symptoms among White adults only. The association between the first experience of parental death during childhood and depressive symptoms was null across all racial-ethnic subgroups.

Social support and strain were significantly associated with early-midlife depressive symptoms across all racial-ethnic subgroups in fully adjusted models: White adults (Cumulative Social Support b = −0.45, 95% CI: −0.53 - −0.38, Cumulative Social Strain: b = 0.60, 95% CI: 0.50 - 0.70), Black adults (Cumulative Social Support b = −0.40, 95% CI: −0.53 - −0.28, Cumulative Social Strain: b = 0.59, 95% CI: 0.39 - 0.80), and Hispanic adults (Cumulative Social Support b = −0.37, 95% CI: −0.52 - −0.22, Cumulative Social Strain: b = 0.71, 95% CI: 0.45 – 0.96). Adjustment for perceived social support and strain reduced the magnitude of the association between parental death on early-midlife depressive symptoms.

Mediation Results

In preparation for the mediation testing, we tested the relationship between our primary independent variable – life course stage of first parental death – to early-midlife social support and strain (Supplementary Tables 10 - 13) (Woodward, Quashie, & Zhang, 2026). In the full analytic sample, parental deaths during young adulthood (b = −0.03, 95% CI: −0.48 - −0.14) predicted decreased cumulative social support in early-midlife, net all social covariates. In comparison, the estimates for all life course stages of first parental deaths were positively, yet not significantly, associated with cumulative strain in early-midlife. Ancillary analyses revealed that the association between first parental death during young adulthood and cumulative social support was primarily driven by trends within the White adult subsample (b = −0.26, 95% CI: −0.41 - 0.11).

The findings from the KHB mediation analysis, Table 4, revealed that cumulative social support and strain collectively account for approximately 18.75% and 22.50% of the association between first parental death in young adulthood to perceived stress and depressive symptoms in early-midlife, respectively. Individually and independently of each other, cumulative social support explains 15.96% (z = 2.03; p < 0.05) and cumulative social strain explains 2.79% (z = 0.30; p > 0.05) of the association between first parental death in young adulthood to early-midlife perceived stress, whereas cumulative social support independently explains 19.57% (z = 2.05; p < 0.05) and cumulative social strain explains 2.93% (z = 0.32; p > 0.05) or depressive symptoms. For both perceived stress and depressive symptoms in early-midlife, cumulative social strain does not statistically significantly mediate the relationship with first parental death in young adulthood. Additionally, ancillary analyses suggest that this pattern of findings was driven by the White adult subsample for perceived stress (Cumulative Social Support % reduction 23.54%, p < 0.01) and depressive symptoms (Cumulative Social Support % reduction 27.39%, p < 0.01).

Discussion

The current study is among the first to examine the complex relationships between the timing of parental death across the life course, quality of social ties, and early-midlife psychological distress across racial-ethnic groups using a nationally representative US sample. The comparative design of this study sought to determine whether the life course stage at first parental death differentially shapes consequential social ties and psychological distress across racial-ethnic identities. Our findings contribute to and extend previous bereavement literature. First and foremost, we found that early-midlife psychological distress was associated with the timing of the first parental death experience in the life course observed exclusively among White individuals, thereby, revealing racial patterning consistent with the "Minority Mental Health Paradox".

Contrary to our expectations, parental death in childhood did not predict an increase in psychological distress outcomes later in early midlife; rather, the loss of a parent during young adulthood and early-midlife proved more potent (Hypothesis 1). More recent theoretical discourse suggests that the psychological impact of parental death is more heterogeneous than traditionally assumed (Bonanno, 2019). Although traditional grief frameworks posit that early experiences of parental death are deterministic of psychological distress throughout the life course (Bowlby, 1979), recent trends in bereavement science suggest that the majority of individuals often report few or no grief or psychological symptoms after a few years following the death (Bonanno, 2019; Pan et al., 2025). The null findings with parental death during childhood might reflect the interplay of internal resilience traits and timely external protective factors that foster healthy coping over time (Bonanno, 2019).

Conversely, the statistically significant associations observed in young adulthood and early-midlife may indicate that parental death has greater psychological consequences during life stages defined by increased independence and responsibility (Umberson, 2003; Kamis, Stolte, & Copeland, 2022). Parental death can disrupt an adult child's generational role, responsibilities, and relationships, thereby generating psychological distress (Umberson, 2003). The experience of parental death during young adulthood may delay personal pursuits or milestones – such as higher education, career advancement, and childbearing – detrimentally influencing long-term adult success (Umberson, 2003; Kamis, Stolte, & Copeland, 2022). Although some of these challenges may be less pertinent in early-midlife, parental bereavement still may have implications for family structure and connectivity, thereby imposing additional burdens on adult children (Umberson, 2003). Considering increasing responsibilities in conjunction with disrupted social ties, adults who have experienced parental death may face elevated levels of psychological distress (Umberson, 2003; Kamis, Stolte, & Copeland, 2022).

Second, our findings highlight the importance of considering racial-ethnic differences when examining the linkages between the life course timing of parental death and psychological distress in early-midlife. Only parentally bereaved White adults experienced elevated perceived stress and depressive symptoms in early-midlife. Therefore, our findings lend support to the "Minority Mental Health Paradox", which suggests that greater exposure to social disadvantage and discrimination is often inversely related to psychological challenges among systematically minoritized communities in the US (Shuey & Willson, 2008). Indeed, there were no statistically significant associations between parental death at any life course stage and either psychological distress outcome in early-midlife for Black adults. Overall, the null findings among minoritized adults likely indicate qualitative differences in parental bereavement experiences across racial-ethnic identities.

Individuals do not experience bereavement in a social vacuum as they live within broader social contexts with other multilevel stressors and resources (Thompson et al., 2016; Wilson & O’Connor, 2022). The manner in which individuals experience stressors, both incidental events (e.g., parental death) and systemic stressors (e.g., discrimination), significantly shapes stress appraisal and response (Pearlin, 2010). Specific to bereavement stress, Black individuals face earlier parental death(s) compared to White individuals, implicating underlying structural disadvantage (Umberson, 2017). Communities with the highest prevalence of parentally bereaved youth are often characterized by other societal challenges, such as lower socioeconomic stability, fewer community resources and assets, and discrimination (Shuey & Willson, 2008). The earlier timing and higher frequency of parental death(s) can further deplete limited resources in said communities (Burns et al., 2020). The combination of systemic disadvantage and potential racial differences in the experience of grief may shape the lack of statistical associations between parental death and psychological distress among early-midlife Black adults (Wilson & O’Connor, 2022).

Our third and fourth hypotheses tested the role of social relationship quality between life course timing of first parental death and psychological distress in early-midlife. Partially supporting Hypotheses 3 and 4, cumulative social support was associated with lower and cumulative strain was associated with higher psychological distress. The main effect of parental death timing on both psychological distress outcomes was attenuated after adding both cumulative social support and strain, driven primarily by the White and Hispanic subsamples (Hypothesis 4). These findings suggest that the quality of social relationships may partially explain the observed associations between parental death and early-midlife psychological distress within these racial-ethnic groups. This association may have diverged for Black adults given previously observed racial differences in relationship quality following parental death, such as with their grown children (Kim et al., 2021). Our findings align with existing work suggesting that the quality of social relationships is influential to early-midlife psychological distress and health (Thomas, 2016; Kim et al, 2024) and that there are meaningful racial differences meriting further exploration (Donnelly et al, 2023).

Finally supporting Hypothesis 5, we found that cumulative social support independently mediated the association and explained between 15 – 20% of the relationship between first parental death during young adulthood and early-midlife psychological distress in the full analytic sample. This finding demonstrates how the consequences of parental death during young adulthood on perceived social support can have broader ripple effects, which in turn, can exacerbate early-midlife psychological distress (Kim et al., 2021). In contrast, cumulative social strain did not statistically significantly mediate the relationship between parental death and early-midlife psychological distress overall. Therefore, negative social interactions following parental death may not lead to higher levels of psychological distress; rather, the lack or absence of positive cumulative support could be a more influential pathway. This finding contrasts with prior work that found negative social interactions had a salient mediating effect on the relationship between traumatic events and psychological distress among White adults (Lincoln et al., 2003). Additionally, lower cumulative social support was only a statistically significant mediator following parental death during young adulthood, suggesting it serves as a distinctly strong pathway during this transitory life stage when individuals are highly vulnerable to the loss of important social resources provided by parents (Umberson & Donnelly, 2023).

Study Limitations & Future Directions

The current study is not without limitations. First, statistical power concerns precluded the inclusion of all racial groups collected in the Add Health study, such as Asians and Native Americans/Indian Americans, which prevents their unique bereavement experiences from being adequately examined in the current study. Second, the Perceived Stress Scale-4 (PSS-4) (Cohen, Kamarck & Mermelstein, 1983), may be interpreted differently across racial-ethnic groups, as previous research has shown that stress appraisal may be lower among Black and Hispanic adults relative to White adults (Brown, Mitchell & Ailshire, 2020). Therefore, the findings may be indicative of measurement artifacts rather than reflecting the true relationship under study. To that end, our third limitation was that the measures for depressive symptoms (CES-D-5) and perceived stress symptoms (PSS-4) were used as proxy coping measures given the unavailability of grief measures in the Add Health study. They are state-specific measures and do not assess formal diagnoses of mental health outcomes, such as depression or anxiety disorders (Laurie & Neimeyer, 2008). Future US nationally representative survey efforts could consider the incorporation of formal grief measures into routine data collection to better understand the nuances of this universal experience. Fourth, this study lacked death contextual information, including the cause of the parental death, which could influence the magnitude and direction of findings. Fifth, this study was unable to ascertain the proximity of social network individuals and other groups of people aside from spouses/partners, children, extended family, and friends over time, and whether support was experienced in other ways, such as through financial and instrumental support. Future bereavement research may benefit from more refined operationalizations of social support and strain capturing detailed relationship histories with a broader social network over time. Lastly, Westernized frameworks (e.g., Life Course Theory and the Minority Mental Health Paradox) tend to homogenize and center the experiences of White individuals as the “standard” and consider divergent experiences as the “exception to the rule.” Future bereavement research should be driven by more context-sensitive theory that respects cultural variability and places greater emphasis on the distinct experiences faced by minoritized groups.

Conclusions

The primary contribution of this work lies in the joint comparison of racial-ethnic identity and life course stage on psychological distress and social ties within the bereavement context. We found that the experience of parental death in young adulthood and early-midlife was more strongly associated with early-midlife psychological distress compared to parental death in childhood. These findings were not observed amongst Black and Hispanic adults, potentially due to differences in the interpretation of parental death as a life stressor and expressions or management of grief compared to White adults. Finally, cumulative social support partially drove the association between parental death in young adulthood and early-midlife psychological distress, uncovering the necessity of social support for adaptive coping during this life stage. These findings encourage further exploration of how bereavement care and community-based services can effectively foster social connection and support among those experiencing parental death in young adulthood and early-midlife (Lichtenthal et al., 2024). This is particularly important as the incidence of premature parental death continues to climb within minoritized communities due to systemic racism and state-sanctioned violence in the US (Wilson & O’Connor, 2022). By understanding the multilevel factors of parental death stress within an integrative sociocultural framework, future programming and policy can be tailored to account for the full extent of disadvantage and provide targeted support that addresses the unique needs of bereaved individuals.

Key messages.

• US Black individuals have a significantly higher prevalence of parental deaths across all life course stages.

• There are racial-ethnic differences in psychological distress following parental death across life course stages that closely align with the “Minority Mental Health Paradox”.

• There were observed differences in perceived midlife social support and strain by life course stage of parental death across racial-ethnic groups.

• KHB mediation analysis showed that cumulative social support in midlife independently accounted for approximately 16% and 20% of the association between parental death in young adulthood and early-midlife stress and depressive symptoms, respectively.

Data availability statement

This research uses data from Add Health, funded by grant P01 HD31921 (Harris) from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD), with cooperative funding from 23 other federal agencies and foundations. Add Health is currently directed by Robert A. Hummer and funded by the National Institute on Aging cooperative agreements U01 AG071448 (Hummer) and U01AG071450 (Aiello and Hummer) at the University of North Carolina at Chapel Hill. Add Health was designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris at the University of North Carolina at Chapel Hill. Requests for the Add Health restricted dataset can be made here: https://addhealth.cpc.unc.edu/data/restricted-use-data-sets/