Aging and Resilience in Vietnamese American Older Adults: Evidence from the Vietnamese Insights into Cognitive Aging Program

Abstract

Objectives:

Vietnamese American older adults have endured early life adversity and war-related trauma exposures, which are associated with worse aging outcomes. However, not all individuals experience poor outcomes, suggesting the possibility of protective factors. We examined associations between early life adversity, war-related exposures, and later life outcomes, and whether later life resilience moderated these associations.

Design:

This was a cross-sectional study.

Setting:

Enrollment occurred at two community-based sites in California.

Participants:

Participants were from the Vietnamese Insights into Cognitive Aging Program (VIP), a longitudinal cohort study of older Vietnamese Americans.

Measurements:

Predictors included war-related exposures, early life adversity, and psychological resilience. Outcome variables were depression symptoms, perceived stress, and global cognition.

Results:

We analyzed data for 511 participants in VIP (mean age = 72.6, SD = 5.1). A jointly-estimated moderation model using Bayesian linear regression revealed conditional associations of wartime evacuation/displacement with greater depression symptoms (β = 0.183, 95% CI [0.085, 0.280]), and exposure to harsh conditions and witnessing death with greater perceived stress at the mean level of resiliency (β = 0.160, 95% CI [0.046, 0.270]). Greater resilience attenuated the strength of association between exposure to harsh conditions and witnessing death with depression symptoms (β = −0.111, 95% CI [−0.200, −0.023]).

Conclusions:

These findings potentially support stress-buffering theories and identify resilience as a possible protective factor in this older Vietnamese American sample. Results underscore the need for longitudinal studies to examine the role of resilience in buffering early life adversity on late life health and well-being.

INTRODUCTION

Early life adversity can negatively impact later life cognition and psychological outcomes [1–3]. Adverse childhood experiences (ACEs) – such as abuse, neglect, and household dysfunction – are linked to poorer cognitive performance in domains such as memory, executive function, and attention in later life among older adults and can leave a lasting negative impact on overall health [4,5]. Older adults with a history of early adversity report higher levels of perceived stress and are more likely to experience recurrent depressive episodes [6].

Although the detrimental effects of early life adversity on cognitive and psychological functioning are well-documented, the outcomes are not universally negative. A growing body of research highlights variability in how individuals respond to adverse experiences, with some demonstrating resilience and achieving relatively favorable physical and mental health outcomes in later life [7]. In the mid-20^th century, the concept of psychological resilience gained traction as researchers sought to understand why some individuals thrived despite facing adversity [8]. Werner’s influential study on children in Hawaii in the 1950s, identified protective factors that enabled some children to grow into thriving adults despite having faced high-risk conditions [9,10]; this laid the foundation for resilience research. Rutter emphasized resilience as “relative resistance to psychosocial risk experiences,” highlighting the interplay of genetic, environmental, and social factors in moderating outcomes [11,12]. In recent decades, resilience has been viewed no longer as a static trait only a few extraordinary individuals possess, but a result of other protective factors that contribute to the development of skills that result in enhanced self-regulation, in part due to environmental and social supports [13]. Today, resilience is viewed as a dynamic process involving interactions between biological, psychological, and social systems that enable individuals to maintain or restore mental health despite facing adversity [14]. In the work of Connor and Davidson, resilience was viewed as a measure of successful stress-coping ability, thus an important target of treatment in anxiety, depression, and stress reactions [15]. We agree with this view and similar to Ong and colleagues’ seminal paper [16], believe resilience is a disposition that can change and evolve over time, as positive adaptation is a core component of resilience. Recovery from trauma is just one phenomenon that existing conceptualizations of resilience have described [17].

Previous work has found that resilience buffers the effects of early life adversity on later life depression, thereby reducing the likelihood of developing depression symptoms and cognitive decline later in life [18,19]. A study conducted with a predominantly African American adult population showed that resilience, viewed as a protective factor against psychological consequences of childhood abuse and other traumatic experiences, mitigated depressive symptom severity as both a main effect and as a moderator of trauma exposure [20]. More recently, Du et al. [21] provided longitudinal evidence from the U.S. Health and Retirement Study showing that psychological resilience and global cognitive function were positively and reciprocally associated over time. This mutual reinforcement suggests that resilience not only plays a direct role in maintaining cognitive health in aging populations, but might protect against emotional distress.

Vietnamese American older adults represent a population that has endured significant early- and mid-life adversity, including war-related trauma, forced migration, and post-resettlement stressors [22]. Many individuals fled Vietnam during or after the Vietnam War, often under life-threatening conditions, and resettled in the U.S. as refugees or immigrants. These experiences have been linked to long-term psychological distress, including elevated rates of depression, anxiety, and post-traumatic stress disorder (PTSD) in later life [23]. A life course perspective suggests that both pre- and post-migration traumas – such as exposure to violence, loss of family, and cultural dislocation – continue to shape mental health trajectories well into older adulthood [23].

Despite these challenges, not all Vietnamese American older adults experience poor health outcomes, suggesting the possible presence of protective factors that may mitigate the long-term effects of early adversity. Although resilience has been studied in various migrant populations, its specific role in buffering the effects of war and immigration trauma among Vietnamese American older adults remains underexplored. In this study, we hypothesized that war exposures and early life adversity are related to poorer later life health in Vietnamese Americans, as measured by perceived stress, global cognition, and depression symptoms (see Figure 1). Furthermore, we hypothesized that these associations are attenuated in Vietnamese American older adults with higher psychological resilience.

Figure 1. Hypothesized model of associations between early life adversity and war exposures on later life outcomes in older Vietnamese Americans.

Early life adversity and war exposures are associated with more stress, lower cognition, and higher depressive symptoms. Higher resilience will attenuate the negative impact of exposures on outcomes (covariates not shown).

Methods

Sample and Design

This cross-sectional study included data from participants enrolled in the ongoing Vietnamese Insights into Cognitive Aging Program (VIP), a longitudinal study of cognitive aging in older Vietnamese Americans in Northern California [22]. Individuals were included if they identified as Vietnamese or Vietnamese American, were at least 65 years of age, spoke Vietnamese or English, and migrated from Vietnam. Respondents were excluded if they had an unstable major medical illness (e.g., active cancer), major primary psychiatric disorder (e.g., schizophrenia), or lived in a nursing home or assisted living facility. Participants could be included in VIP if they had cognitive impairment as long as they could complete the majority of the study visit. At the start of the baseline/initial visit, written informed consent was obtained. The study protocol for the VIP was approved by the institutional review boards at the two recruitment sites, the University of California Davis and University of California San Francisco. Baseline visits began in January 2022 and were completed in November 2023. The entire baseline visit lasted approximately 3–3.5 hours, while the two follow-up assessments (conducted approximately annually after the previous assessment) lasted about 2–3 hours. All assessments were conducted in the Vietnamese language [24].

Measures

Demographic data.

Age, gender, marital status, and educational level were included as covariates in this study (Table 1). Age was treated as a continuous variable. Gender was dichotomized as male (0) and female (1). Marital status was dichotomized as married/living with partner (1) versus separated/widowed/divorce/single (0); educational level was categorized into six groups: didn’t go to school (0), primary school (grades 0–5=1), middle school (grades 6–9=2), high school (grades 10–12=3), some college/associate degree (=4), and college/graduate degree (=5).

Table 1.

Demographic characteristics of participants in the VIP Study (N=511).

	n (%)	Mean	SD	Range	% Missing
Age (years)		72.58	5.13	65 – 89	0.0%
Gender
Male	230 (45.0%)				0.0%
Female	281 (55.0%)				0.0%
Marital status
Married/living with partner	311 (60.9%)				0.0%
Separated/widowed/divorced/single	200 (39.1%)				0.0%
Education level
Didn’t go to school	7 (1.4%)				0.0%
Up to primary school (grades 1–5)	64 (12.5%)				0.0%
Middle school (grades 6–9)	70 (13.7%)				0.0%
High school (grades 10–12)	179 (35.0%)				0.0%
Some college/Associate degree	123 (24.1%)				0.0%
College/Graduate degree	68 (13.3%)				0.0%
English speaking proficiency
Very well/Well	96 (18.8%)				0.0%
Some/A little bit	366 (71.6 %)				0.0%
Not at all	49 (9.6 %)				0.0%
Years lived in the U.S.		26.8	13.0	0–56	0.0%
Perceived Stress		14.89	6.35	0 – 32	0.2%
Depression symptoms		1.63	0.49	1 – 3.54	0.2%
Resilience		25.48	6.91	0 – 40	0.4%
Harsh conditions and witnessing death (RC1)		0.015	0.80	−1.43 – 3.36	0.0%
Personal life threat/endangerment (RC2)		−0.002	0.68	−2.69 – 2.65	0.0%
Family endangerment (RC3)		−0.005	0.76	−2.10 – 2.91	0.0%
War-time evacuation/displacement (RC4)		0.005	0.95	−2.28 – 3.27	0.0%
Early life adversity composite		0.099	0.01	0.00 – 0.40	0.2%

Independent variables in the study were war-related trauma exposures (occurring as a result of war) and early life adversity.

War-related trauma exposures included 17 items adapted from the War-Related Traumatic Events Scale (WTES) [25], as modified by the Vietnam Health and Aging Study [26] using elements from the National Vietnam Veterans Readjustment Study [27], Deployment Risk and Resilience Inventory [28], and Combat Exposure Scale [29], were analyzed using principal component analyses with multiple imputed data (details in Meyer et al.[30]). This process yielded four distinct components representing individual level war-related exposures: RC1: harsh conditions and witnessing death; RC2: personal life threat/endangerment; RC3: family endangerment; and RC4: war-time evacuation/displacement (Supplementary Table 1).

Early life adversity was measured using items from the Life Experience and Activities Form (LEAF), which characterizes experiences across the participant’s life span. Following prior studies [3,31], early life adversity was defined as a composite variable constructed as the mean of the following binary responses: mother losing children before or during childbirth (yes/no), number of siblings who died before age 18 (modified to any sibling who died before age 18, yes/no), household member with substance abuse problem (yes/no), household member who had a mental illness (yes/no), and household member who was sent to jail, prison, or re-education camp (yes/no; item unique to the Vietnam War) [22].

The moderator examined was psychological resilience, assessed by the 10-item Connor-Davidson Resilience Scale (CD-RISC). This study utilized the existing translated and validated version of the 10-item CD-RISC that, showed good internal consistency, construct validity and psychometric properties among Vietnamese young adults [32]. It also has been administered in a sample of older Vietnamese caregivers [33] as well as in older Chinese migrants [34]. Sample items include, “I am able to adapt when changes occur”, and “I am able to handle unpleasant or painful feelings like sadness, fear, and anger.” Items were rated on a 5-point Likert scale, ranging from 0 – 4, or from “not true at all” to “true nearly all the time”. The total scale score (range: 0 – 40) was used, with higher scores indicating greater resilience and ability to cope with challenges that arise.

Dependent variables were psychological and cognitive outcomes which included perceived stress, depression symptoms, and global cognition.

Perceived stress was measured by the Perceived Stress Scale (PSS-10) which assesses an individual’s perceived stress level on a 0 – 4 scale across all items, ranging from “never” to “very often” [35]. The scale has acceptable to high internal consistency, with Cronbach’s alpha values ranging from .78 to .91 [36]. We used the average score across the items of the scale as the outcome variable.

Depression symptoms were assessed by the Hopkins Symptom Checklist-15 (HSCL-15), a self-report measure that assessed depression symptoms over the previous week [37]. The HSCL-15 has good validity and reliability, with Cronbach’s alphas of 0.89–0.90 [38,39], and its structure is generally viewed as unidimensional [40]. In the VIP, two items were removed, including loss of sexual interest or pleasure (following the practice of Hinton et al., 1994) and thoughts of ending life, due to cultural considerations and the absence of a clinician on site to ensure appropriate clinical support. This adjustment resulted in 13 items rated on a Likert scale from 1 to 4. Following the approach in previous studies [38,41], we used the average score of the items. In this study, the 13-item scale had a good internal reliability of α=0.90.

Global cognition was assessed using a previously developed global cognitive composite score that was harmonized with the National Alzheimer’s Coordinating Center Uniform Data Set version 3 neuropsychological test battery [42]. This composite score, which was developed using item response theory, incorporates measures of memory, auditory attention and working memory, visual attention and executive function, and language using tests that were designed for cross-cultural application and were translated and used in the Vietnamese population in the U.S [43]. These global cognition factor scores were estimated with person-specific standard errors; this measurement error was accounted for in our analyses.

Data analysis

Participants’ characteristics are described using means and standard deviations (SDs) for continuous variables and counts (percentages) for categorical variables. A jointly-estimated moderation model was performed using Bayesian linear regression with the brms package (version 2.23.0) [44] for R (version 4.5.2) [45] to simultaneously examine the direct associations between the four war exposure components and the early life adversity composite and depression symptoms, stress, and global cognition, as moderated by resilience (modeled using interaction terms). Age, gender, education, and marital status were included as covariates, with education treated as an ordered categorical variable modeled using monotonic effects. Predictors of resilience were estimated from the same model. Model estimation was performed using Hamiltonian Monte Carlo with 8,000 sampling iterations (4,000 of which were warmup draws), with weakly informative priors (𝒩[0, 1]) on the regression coefficients. Missing data were handled via full information Bayesian imputation, and residual covariances were freely estimated across the three different outcomes modeled simultaneously. To interpret parameter estimates, we estimated and reported standardized coefficient estimates with 95% credible intervals (95% CI). Out of concern for overadjustment, we did not include covariates that may be consequences of early life adversity and war exposures (e.g., PTSD), and thus potential mediators. See Figure 1 for a path diagram depicting the analytic framework [46].

Results

Sample Descriptives

VIP participants (N = 511) ranged in age from 65 to 89 years (Mean = 72.6, SD = 5.1; Table 1). The majority were female (55%) and married or living with a partner (61%). Participants had lived in the U.S. for an average of 26.8 ± 13 years (range: 0 – 56). The majority spoke “some” or “a little bit” of English (72%). Level of highest education varied: about 14% had up to a grade 5 education, 14% middle school, 35% high school, 24.1% some college or an associate degree, and 13.3% college or a graduate degree.

Regression Analyses

Predictors of Resilience

Based on the results shown in Table 2, higher educational attainment (β = 0.132, 95% CI [0.069, 0.216]) and being married (β = 0.262, 95% CI [0.064, 0.456]) were both associated with higher self-reported resilience. Other variables, including early life adversity and war exposures, did not strongly predict late life resilience.

Table 2.

Predictors of Resilience.

Predictor	Estimate	SE	95% CI
Intercept	−0.538*	0.193	[−0.97, −0.198]
Age	−0.004	0.011	[−0.025, 0.017]
Female Gender	0.027	0.148	[−0.264, 0.319]
Education	0.132*	0.037	[0.069, 0.216]
Married	0.262*	0.100	[0.064, 0.456]
RC1	0.059	0.046	[−0.033, 0.15]
RC2	−0.016	0.054	[−0.122, 0.09]
RC3	0.029	0.044	[−0.059, 0.117]
RC4	0.052	0.044	[−0.035, 0.138]
Early life adversity	0.021	0.472	[−0.922, 0.969]

Note. RC1 = Harsh conditions and witnessing death; RC2 = Personal life threat/endangerment; RC3 = Family endangerment; RC4 = Wartime evacuation/displacement.

^*95% credible intervals do not overlap 0.

Conditional Associations

Table 3 summarizes the conditional associations (i.e., at the sample mean level of resilience) for the exposures of interest and covariates with each outcome. Wartime evacuation or displacement (RC4) was associated with more depression symptoms (β = 0.183, 95% CI [0.085, 0.280]). In addition, harsh conditions and witnessing death (RC1) was associated with greater perceived stress (β = 0.160, 95% CI [0.046, 0.270]). Global cognition was significantly associated with age (β = −0.042, 95% CI [−0.059, −0.024]) and education (β = 0.494, 95% CI [0.332, 0.643]), indicating that older age was associated with lower cognitive scores and more education was associated with higher cognitive scores. There was less certainty in the association between early life adversity or war trauma and global cognition (i.e., 95% CI contained 0).

Table 3.

Conditional associations between war exposures, early life adversity, and resilience on depression symptoms, stress, and global cognition.

Predictor	Estimate	SE	95% CI
Depression Symptoms
Intercept	0.090	0.274	[−0.473, 0.626]
Age	0.013	0.011	[−0.008, 0.035]
Female Gender	0.058	0.152	[−0.243, 0.360]
Education	−0.015	0.060	[−0.134, 0.107]
Married	−0.181	0.139	[−0.454, 0.096]
RC1b	0.213*	0.052	[0.111, 0.313]
RC2c	0.069	0.056	[−0.041, 0.180]
RC3d	0.053	0.048	[−0.040, 0.148]
RC4e	0.183*	0.050	[0.085, 0.280]
Early life adversity	0.025	0.484	[−0.947, 0.979]
Resilience	−0.267	0.358	[−0.972, 0.448]
Perceived Stress
Intercept	−0.196	0.295	[−0.757, 0.418]
Age	−0.003	0.012	[−0.026, 0.020]
Female Gender	−0.069	0.164	[−0.400, 0.255]
Education	0.081	0.065	[−0.057, 0.203]
Married	0.019	0.157	[−0.302, 0.323]
RC1b	0.160*	0.057	[0.046, 0.270]
RC2c	−0.031	0.059	[−0.147, 0.084]
RC3d	0.072	0.051	[−0.028, 0.170]
RC4e	0.097	0.054	[−0.014, 0.201]
Early life adversity	0.003	0.502	[−1.001, 1.012]
Resilience	0.248	0.420	[−0.608, 1.058]
Global Cognition
Intercept	−1.569*	0.366	[−2.265, −0.815]
Age	−0.042*	0.009	[−0.059, −0.024]
Female Gender	−0.222	0.124	[−0.464, 0.025]
Education	0.494*	0.078	[0.332, 0.643]
Married	−0.001	0.126	[−0.262, 0.242]
RC1b	−0.078	0.045	[−0.170, 0.008]
RC2c	−0.023	0.047	[−0.117, 0.069]
RC3d	−0.054	0.040	[−0.134, 0.024]
RC4e	−0.077	0.042	[−0.164, 0.005]
Early life adversity	0.022	0.363	[−0.733, 0.756]
Resilience	0.212	0.353	[−0.453, 0.967]

Note. Conditional associations are estimated at the sample mean level of resilience.

RC1 = Harsh conditions and witnessing death; RC2 = Personal life threat/endangerment;

RC3 = Family endangerment; RC4 = Wartime evacuation/displacement.

^*95% credible intervals do not overlap 0.

Moderating Effects of Resilience

Table 4 provides the estimated interactions representing the degree to which associations between exposures and outcomes depend on resilience. Results reveal that the association between harsh conditions and witnessing death (RC1) and depression symptoms depended on resilience (β = −0.111, 95% CI [−0.200, −0.023]). As seen in Figure 2, higher resilience was associated with an attenuation of the strength with which harsh conditions and witnessing death was associated with depression symptoms. No other associations were moderated by resilience.

Table 4.

Moderation of exposures’ associations with depression symptoms, stress, and global cognition by resilience.

Exposure	Moderation Effect Estimate	SE	95% CI
Depression Symptoms
RC1	−0.111*	0.045	[−0.200, −0.023]
RC2	−0.075	0.043	[−0.160, 0.009]
RC3	0.060	0.041	[−0.019, 0.141]
RC4	0.039	0.044	[−0.048, 0.125]
Early Life Adversity	−0.010	0.039	[−0.087, 0.067]
Perceived Stress
RC1	−0.055	0.047	[−0.146, 0.037]
RC2	−0.060	0.044	[−0.146, 0.028]
RC3	0.062	0.042	[−0.020, 0.142]
RC4	0.047	0.046	[−0.043, 0.137]
Early Life Adversity	0.002	0.041	[−0.078, 0.082]
Global Cognition
RC1	0.013	0.037	[−0.060, 0.085]
RC2	−0.005	0.035	[−0.074, 0.064]
RC3	0.013	0.033	[−0.052, 0.079]
RC4	−0.021	0.036	[−0.091, 0.048]
Early Life Adversity	−0.058	0.032	[−0.121, 0.005]

Note. RC1 = Harsh conditions and witnessing death; RC2 = Personal life threat/endangerment; RC3 = Family endangerment; RC4 = Wartime evacuation/displacement.

^*95% credible intervals do not overlap 0.

Figure 2.

Depression symptom severity as a function of RC1 (exposure to harsh conditions and witnessing death; x-axis) and resilience (mean and ±1 SD levels) estimated by the Bayesian regression model. Values in the plot refer to a hypothetical reference participant, who is a 70-year-old unmarried male with 10–12 years of education and sample mean values for other exposures (RC2, RC3, RC4, and Early Life Adversity).

DISCUSSION

This study examined the cross-sectional associations between early life adversity and war exposures with later life outcomes, specifically perceived stress, depression symptoms, and global cognition, jointly estimated in a cohort of Vietnamese American older adults. We found that out of the five war-related exposure and early life composite components examined, two exposures (harsh conditions and witnessing death, wartime evacuation and displacement) were associated with later life outcomes. In particular, both exposures were associated with depression symptoms, but only harsh conditions and witnessing death was associated with perceived stress. Additionally, we found that the degree to which depression symptoms were associated with harsh conditions and witnessing death was dependent upon resilience. That is, the association between exposure to harsh conditions and witnessing death and depression symptoms was attenuated for individuals with higher levels of resilience. These findings are aligned with the existing literature on childhood abuse and traumatic early life experiences and resilience [4,6,20,47].

In addition, this study found that education and marital status were associated with self-reported resilience. Specifically, married participants and those with more education reported greater levels of resilience, independent of age, gender, education, and adversity and war exposures. These results mirror other studies showing a similar association of marital status [48,49] and education [50,51] with higher resilience and/or better mental health. Being married and having more education may contribute to greater resilience via mediating factors such as social support or greater access to health-promoting and socioeconomic resources. Although we did not find that war exposures and early life adversity were associated with global cognition at the mean level of resiliency, the effect size and direction of estimates for the war exposures suggested the potential for a small negative impact on cognition that could not be reliably distinguished from a null or positive effect with more than 95% probability. It is possible that the effects of trauma on cognition may not be specific to any one source of trauma, but could instead be related to the overall trauma burden. Alternatively, a recently published paper indicated that war exposures and early life adversity were associated, not with global cognition, but with domain-specific measures of cognition in nuanced ways (e.g., executive function, language, visuospatial skills)[52].

Many Vietnamese American older adults in this study lived through decades of war, including the Vietnam War, and were exposed to harsh conditions such as the inability to sleep due to bombings and displacement, severe food and clean water shortages, imprisonment or forced labor in re-education camps, exposure to toxic chemicals, and witnessing death and injury of civilians and soldiers [3,53]. These experiences constitute extreme and prolonged traumatic stressors that are known to have long-term psychological consequences [54]. In this study, these exposures to harsh conditions and witnessing death were associated with depression symptoms in later life. However, many Vietnamese American older adults demonstrated psychological resilience and hardiness [55,56]. This resilience may play an important role in later life outcomes.

Resilience can potentially mitigate the effects of trauma through several psychological and neurobiological mechanisms, such as cognitive reappraisal and adaptive coping, neurobiological regulation, or social support and the ability to construct meaning from adversity. People with higher resilience tend to engage in positive cognitive reframing, which can help reduce the emotional impact of traumatic memories [57]. Having higher resilience is associated with lower cortisol reactivity, which helps regulate stress responses and prevent prolonged depressive episodes [58]. Furthermore, resilient individuals may be more likely to seek social support and engage in meaning-making, potentially reducing feelings of helplessness and despair following trauma exposure [23].

Implications

Results of the present study extend the body of research on resilience by focusing on Vietnamese American older adults, a group affected by war and (forced) migration-related trauma but whose resilience mechanisms remain understudied. Findings from this study provide empirical support for resilience playing a potentially protective role in early life exposures and support the themes of adversity and strength recently published in a qualitative study of VIP participants [53]. For Vietnamese American older adults with a history of war and displacement, resilience may be a protective factor that enhances health and mental health in later life. These findings may inform resilience-building strategies for aging and mental health interventions for trauma-exposed immigrant populations. Specifically, among Vietnamese American older adults, culturally tailored, resilience-building strategies may be particularly helpful in mitigating the lasting effects of war-related trauma and forced migration [23].

Limitations and Strengths

Although this study provided valuable insights, some limitations should be acknowledged. First, reliance on self-report measures may introduce recall bias, especially for events that occurred decades in the past [59]. Second, the cross-sectional design of this study limits the ability to assess temporal dynamics and raises the possibility of reverse causality, making it difficult to disentangle the directionality between resilience and depression symptoms (e.g., depression symptoms may influence self-reported resilience). However, future VIP research will employ longitudinal designs to examine how resilience evolves over time and how it may buffer against the effects of ongoing stressors in this population. We also will examine if variables, such as PTSD, might be a potential mediator of early life exposures on later life outcomes. Third, resilience was measured decades after the trauma, raising questions about the stability of psychological resilience over the course of life, although this is slightly assuaged by the literature on both trait and process conceptualizations of resilience in older adults [16]. Fourth, removing the two items from the Hopkins depression symptoms measure, while we felt was culturally and ethically reasonable, may have affected the validity of the measure. Fifth, we did not adjust for multiple comparisons across exposures and outcomes leaving the possibility of detecting an association or interaction due to chance. Lastly, we may be underpowered to detect subtle moderating effects of resilience, despite the large sample size.

Other than education, which had a moderate effect size in terms of its association with cognition, other exposures had small effect sizes. However, these effects are similar to what has been reported in other studies of early life exposures and later life health [60]. Considering there are many factors that contribute to cognitive aging and dementia risk (e.g., education, vascular factors), understanding how exposures that happened decades ago, yet still have an impact on older adult health, could be important for prevention and intervention. Moreover, known modifiable lifestyle factors for dementia that together account for 45% of dementia cases, also individually have small effect sizes[61]. Thus, despite the magnitude of effect sizes in this study, there still are important implications for cognitive health.

This study also had several notable strengths. First, it stands as one of the few empirical investigations that specifically examines how early life exposures are associated with later life cognitive and psychological outcomes among a large sample of Vietnamese American older adults. Despite the distinct historical context marked by war, forced migration, and resettlement, this population is largely underrepresented in aging and mental health research. By examining how these early life exposures affect later life health in an immigrant/refugee population, this study addresses a critical gap in the literature. Second, the use of Bayesian regression modeling was a methodological strength. It enabled simultaneous estimation of conditional associations and interactions, accounted for measurement error, and accommodated complex relationships between study variables. Lastly, the administration of the exposure and outcome assessments in the Vietnamese language – including a validated and comprehensive measure of global cognitive functioning – is a methodological strength as all measures went through extensive pilot testing [24]. It will be important for future work in this population to examine how these assessments change over time.

In conclusion, the present cross-sectional study underscores the potentially protective role of resilience in mitigating the long-term effects of exposure to harsh conditions and witnessing death in early life on psychological outcomes among Vietnamese American older adults. These results can inform culturally sensitive mental health strategies to support healthy aging in an understudied population. The findings require further replication in larger studies and extension into longitudinal designs, which, all together, could be used to inform policy and interventions that enhance resilience and promote psychological well-being in other populations exposed to a history of war and conflict.