Adverse Childhood Experiences, Psychological Distress, and Resilience in Health Professions Students

Abstract

Purpose

To determine the relationship between adverse childhood experiences (ACEs), social disadvantage, psychological distress, and resilience in graduate health professions students.

Method

This study includes cross-sectional analyses from a longitudinal survey of medical, veterinary, and advanced practice provider students at matriculation to the University of California Davis in July 2019. The survey contained an expanded Adverse Childhood Experiences Questionnaire (ACEs-14), a measure of psychological distress (the Medical Student Well-Being Index [MSWBI]), and the Brief Resilience Scale. Responses were linked to demographics, including markers of social disadvantage (female gender, underrepresented in medicine [URM] status, and first-generation college graduate [first-gen] status). The relationships between ACEs, social disadvantage, psychological distress, and resilience were tested using linear or logistic regression.

Results

Complete survey responses were provided from 240 of 357 students (67% completion rate). About two-thirds of students (67%, 161/240) reported ≥ 1 ACE, while a quarter (25%, 60/240) reported ≥ 4 ACEs. URM and first-gen students had higher odds of reporting ≥ 4 ACEs (odds ratio [OR] = 1.56; P = .049 and OR = 2.63; P < .001, respectively) than their nondisadvantaged peers based on binary logistic regression analysis. Higher ACEs-14 scores were associated with higher psychological distress scores (P < .001). The majority of students reported normal or high resilience (normal: 76%, 183/240; high: 10%, 25/240) regardless of ACEs-14 scores. There was not a statistically significant relationship between ACEs-14 scores and resilience scores (P = 0.936).

Conclusions

Health professions students from some socially disadvantaged backgrounds at this institution reported statistically significantly higher ACEs-14 scores than their nondisadvantaged peers. Childhood adversity was associated with increased psychological distress but not with low resilience. Implications for equity- and trauma-informed health professions education and interventions are discussed.

Adverse childhood experiences (ACEs), such as child maltreatment and witnessing family violence, are associated with negative health and psychosocial outcomes.¹ Research has shown that trauma exposure as measured by ACEs fosters cycles of adversity, disadvantage, and poor health in families and communities, increasing the risk of similar trauma exposures in subsequent generations.² Also, ACEs tend to co-occur,³ have a measurable impact on health even decades later,^1,4 and have health-harming effects that are cumulative.⁵

After establishing a causal link between trauma exposure and mental disorders,⁶ more recent research has focused on the prevalence of ACEs on health professions students and their effect on psychological distress.^7–13 Health professions students are those pursuing graduate degrees in the health professions, such as medical, dental, nursing, pharmacy, veterinary, social work, and related professions (hereafter referred to as students). Practicing health care professionals and students are in a unique position to provide trauma-informed care and address trauma-related health problems across the lifespan.¹⁴ At the same time, they are at increased risk of psychological distress^15–20 and secondary traumatic stress.^21–24 As such, understanding the prevalence and impact of ACEs on these students is important for improving educational and professional outcomes, as well as for improving the care of their future patients.¹⁴

Although research shows that female gender, low socioeconomic status (SES), and intersecting socially disadvantaged identities are associated with increased prevalence of ACEs,^25–29 one notable gap in ACEs research is whether a socially disadvantaged background may intensify the negative influence of ACEs on students’ mental health. Addressing this gap is important as health professions schools invest considerable resources to increase the number of individuals from groups that are historically underrepresented in medicine (URM; see below for definition) and to decrease or prevent student mistreatment, which is reported more frequently by individuals from these sociodemographic groups.^30–32 Research shows that low SES, being a first-generation college graduate (first-gen), and URM status are negative predictors of matriculating into or graduating from medical and nursing school, as well as transitioning into advanced practice in these fields.^33–37 Evidence of the impact of ACEs on female, first-gen, or URM students may provide benchmarks with which to assess the effect of institutional resources and interventions aimed at increasing student recruitment, retention, and academic progression.

Research focused on strengths shows that resilience may lessen the deleterious effects of ACEs on health well into adulthood.³⁸ Psychological resilience has been defined as an active adaptation to stress resulting from biopsychosocial and ecological interactions³⁹ (as opposed to an inherent, stable personality trait that can be fostered at any time during the lifespan⁴⁰). Hence, numerous interventions to promote resilience in health care workers and students have been published.^41–45 The ultimate goal of our research group is to develop equity- and trauma-informed interventions that consider strengths in addition to vulnerabilities, such as ACEs exposure, to enhance resilience in diverse students. Building on prior work,⁹ this study aims to test the following hypotheses:

1. ACEs will be present in graduate health professions students, especially those from socially disadvantaged backgrounds (i.e., those who identify as female, first-gen, or URM).

2. High ACEs exposure will increase psychological distress for students.

3. High ACEs exposure will decrease resilience for students.

Method

Design

A quantitative, longitudinal survey was developed, based on a review of the literature, by a panel of diverse individuals that included students as well as faculty and administrators with expertise in psychiatry, student wellness, program evaluation, and health professions education. This confidential survey included psychosocial measures related to resilience and well-being and was intended to follow a cohort of students from matriculation to graduation. The letter of invitation to participate included links to health support services, which were also included at the end of each survey. This article reports a cross-sectional analysis from the matriculation time point of this longitudinal study. Student focus groups were consulted during survey development and care was taken to be sensitive to possible triggering language. This study was approved by the University of California Davis Institutional Review Board.

Participants

Study participants were students from several graduate health professions programs at the University of California Davis, including medical, veterinary, and advanced practice provider (APP) students. APP students included nurse practitioner and physician associate students from a single, combined advanced practice health professions program at the school of nursing. An administrator from each school emailed the survey to matriculating students via Qualtrics (Qualtrics, Provo, Utah) in July 2019. Students received a small incentive gift card ($25) for participation. APP students matriculated into 2-year master-level programs, while medical and veterinary medical students matriculated into 4-year doctoral-level programs. A data analyst who was not involved in teaching or assessment retrieved the data via Qualtrics and deidentified the data for the research team. The researchers only had access to the deidentified survey responses.

Measures

Demographic indicators of socially disadvantaged background.

Each schools’ registrar granted the data analyst access to students’ information regarding gender, first-gen status, and race/ethnicity obtained by self-report at application. Given the heterogeneity with which each school grouped race/ethnicity data, we used the Association of American Medical Colleges’ definition of URM (i.e., American Indian or Alaska Native; Black or African American; Hispanic, Latino, or of Spanish Origin; or Native Hawaiian or Other Pacific Islander either alone or in combination with any other race/ethnicity)⁴⁶ to create a URM variable defined as a self-reported race/ethnicity other than White or Asian.

Revised Adverse Childhood Experience Questionnaire (ACEs-14).

The ACEs-14 was developed to improve the original Adverse Childhood Experiences Study scale’s ability to predict health outcomes¹ by adding items to the original 10-item scale that assess peer victimization, peer isolation or rejection, exposure to community violence, and low SES.⁴⁷ The items can be answered “yes” or “no,” and each yes counts as 1 point, with the total score being the sum of the yes answers (range: 0–14). In a nationwide sample of 1,949 children from 10 to 17 years old, the effect sizes of these new adversity items obtained through regression analyses were larger than most of the effect sizes of the original ACE items.⁴⁷ For example, additional items in the ACES-14 increased R² values relative to the original ACEs-10 items from 0.32 to 0.46, which was highly significant (P < .001).⁴⁷ ACEs-14 scores above 4 have been shown in multiple studies to predict negative health outcomes. For example, in an international study of over 10,000 adults using an ACEs cutoff score of 4 or more showed that patients with chronic disease were almost 3 times as likely to be in the “high ACEs” group.⁴⁸

Brief Resilience Scale (BRS).

The BRS is a 6-item instrument that measures resilience as the ability to recover from stress.⁴⁹ In its initial validation, the internal consistency was deemed good (Cronbach’s alpha 0.80–0.91).⁴⁹ Also, the psychometric properties of test-retest reliability, convergent validity, and discriminant predictive validity were good.⁴⁹ The scale uses a 5-point Likert scale for statements such as “I tend to bounce back quickly after hard times.” Responses varying from 1 to 5 (where 1 = strongly disagree and 5 = strongly agree) for all 6 items are added together and the sum is divided by the total number of questions answered, giving an overall score range of 1 to 5. The scores are stratified as 1.00–2.99 being low resilience, 3.00–4.30 being normal resilience, and 4.31–5.00 being high resilience.⁵⁰

Medical Student Well-Being Index (MSWBI).

The MSWBI is a 7-item self-report instrument that measures common dimensions of psychological distress experienced by medical students.⁵¹ Additionally, it can help identify students whose scores place them at risk for mental health problems.⁵¹ To obtain a score, each “yes” answer is given 1 point, and the total score from all 7 responses is added up (range: 0–7). Its content validity index was reported at 0.94 and 0.91 for relevance and representativeness, respectively,⁵¹ with a Cronbach’s alpha of 0.68 for internal consistency.⁵¹ MSWBI scores ≥ 4 have been shown to identify students with low mental quality of life or recent suicidal ideation and thoughts of dropping out, with sensitivity and specificity ≥ 90%.⁵² In this study, we refer to MSWBI scores ≥ 4 as “poor well-being” or “high levels of psychological distress.” The MSWBI has been used in a prospective study with a large cohort of U.S. medical students to identify structural and interpersonal characteristics of medical schools (e.g., lack of peer mentorship) in addition to intrapersonal factors (e.g., non-male gender) associated with lower well-being scores.⁵³ For our study, the language of the MSWBI was adapted to apply to veterinary and APP students as well as medical students by using “professional school” rather than “medical school.”

Analytic strategy

We analyzed the data using SAS version 9.3 for Windows (SAS Institute, Inc., Cary, North Carolina). Comparison of ACEs-14, psychological distress, and resilience scores between categorical group variables were conducted. First, A Shapiro-Wilk test was used to determine whether the data fit a normal distribution. The assumption of normal distribution of data was not met, and thus nonparametric tests were used to compare median values across groups, allowing for a more accurate and conservative analysis of the dataset. Bivariate analyses were performed using a Mann Whitney test to compare median ACEs-14 scores for each category of social disadvantage: (1) URM versus non-URM, (2) female versus male, and (3) first-gen versus continuing generation college graduate (continuing gen). A Kruskal-Wallis test was performed to compare median ACEs-14 scores across the 3 programs (medical, veterinary, and APP) followed by bivariate analyses using a Mann Whitney test. Where relevant, descriptive statistics are reported including medians, interquartile ranges (IQRs), means, and standard deviations (SDs). A P value less than .05 was reported as statistically significant, and relevant P values are reported below for each comparison tested.

Odds ratios for high ACEs score (defined as ACEs-14 score ≥ 4) were also calculated using the binary logistic regression model in SAS version 9.3 for Windows with high ACEs-14 (coded as 1 = ACEs-14 score ≥ 4 or 0 = ACEs-14 score < 4) as the dependent variable, and URM status, gender, or first-gen status as the independent variable. We did not control the false discovery rate in this study as all models were selected (planned) a priori and were all part of a comprehensive, complementary examination of associations.

Associations between resilience and psychological distress scores with ACEs scores and with demographic variables were examined by performing linear regression analyses using the generalized linear model (GLM) procedure in SAS. P values less than .05 were reported as statistically significant. The beta coefficients and standard error of beta coefficients are also reported to describe the magnitude and direction of statistically significant relationships. The association between ACEs-14 and MSWBI was tested using linear regression with MSWBI score as the dependent variable. The association between ACEs-14 and BRS was tested using linear regression with BRS score as the dependent variable. A logistic regression was also performed using the GLM procedure in SAS to further investigate the association between categories of ACEs-14 scores, defined as 0, 1–3, and ≥ 4, where ≥ 4 is considered clinically significant trauma exposure.⁴⁸ The logistic regression tested the association of categories of ACEs-14 scores with either MSWBI or BRS score. Odds ratios for high psychological distress (defined as MSWBI score ≥ 4)⁵² were also calculated using the binary logistic regression model in SAS version 9.3 for Windows with MSWBI ≥ 4 (coded as 1 = MSWBI score ≥ 4 or 0 = MSWBI score < 4) as the dependent variable, and total ACEs-14 score as the independent variable. Finally, a full regression model was used to test the association of ACEs-14 score, BRS score, URM status, gender, and first-gen status on MSWBI score.

To determine whether resilience mediated the effect of ACEs-14 score on MSWBI score, a mediation analysis was performed using the GLM procedure in SAS to test the association between ACEs-14 and MSWBI scores, while controlling for resilience. We also tested for interaction or moderation of the effect of ACEs-14 score on MSWBI score by participant demographics (i.e., gender, first-gen status, and URM status) to further examine effects based on social disadvantage and intersectionality of social disadvantages. As above, P values less than .05 were reported as statistically significant, along with the magnitude and direction of beta coefficients to describe the effects of ACEs-14 score on MSWBI score before and after the addition of each mediating or moderating variable to the model.

Results

A total of 357 students were sent the survey at matriculation, and 261 responded and consented to the study (73% response rate). For the current data analysis, 240 students provided complete survey responses (67% completion rate). The traditional binary gender classification was used because this was how gender data were reported on admission. The sample’s demographics are given in Supplemental Digital Appendix 1 (at [LWW INSERT LINK]).

ACEs prevalence

First, we analyzed the prevalence of trauma exposure by looking at participants’ overall ACEs-14 scores and individual ACEs-14 items. About two-thirds of students (67%, 161/240) reported ≥ 1 ACEs, while 33% (79/240) reported no ACEs. One quarter of students (25%, 60/240) reported ≥ 4 ACEs, with most students reporting 1–3 ACEs (42%, 101/240). No students reported exposure to all 14 ACEs items.

Across student groups, the ACEs-14 items that were most commonly reported were household mental illness (31%, 74/240), peer isolation or rejection (29%, 69/240), low SES (29%, 69/240), parental separation or divorce (24%, 57/240), household substance abuse (22%, 52/240), and emotional abuse (21%, 51/240; see Table 1).

Table 1.

Prevalence of ACEs-14 Items for Medical, Veterinary, and Advanced Practice Provider Students (n = 240), University of California Davis, July 2019

ACEs-14 item	Yes responses, no. (%)
Abuse
Emotional abuse	51 (21)
Physical abuse	19 (8)
Sexual abuse	3 (1)
Neglect
Emotional neglect	42 (18)
Physical neglect	17 (7)
Household challenges
Parental separation or divorce	57 (24)
Parent treated violently	20 (8)
Household substance abuse	52 (22)
Household mental illness	74 (31)
Incarcerated household member	16 (7)
Additional adversities
Peer victimization	27 (11)
Peer isolation or rejection	69 (29)
Exposure to community violence	38 (16)
Low socioeconomic status	69 (29)

Abbreviation: ACEs-14, Adverse Childhood Experience Questionnaire.

ACEs-14 scores were also compared across the health professions programs (medical, veterinary, and APP; see Figure 1). Across the 3 groups, there was a statistically significant difference in ACEs-14 scores (medical: median = 2.0 [IQR = 0.0, 4.0], mean = 2.7 [SD = 2.90]; veterinary: median = 1.0 [IQR = 0.0, 3.0], mean = 1.49 [SD = 1.75]; APP: median = 2.0 [IQR = 0.0, 4.0], mean = 3.04 [SD = 3.47]; P = .0123). Medical students had statistically significantly higher average ACEs-14 scores than veterinary students (medical vs. veterinary: P = .007; APP vs. veterinary: P = .021). Medical and APP students did not have significantly different ACEs-14 scores (P = .797).

Figure 1.

ACEs-14 scores across demographic groups (n = 240), University of California Davis, July 2019. The distributions of ACEs-14 scores are depicted using a box and whisker plot for 3 comparison groups of social disadvantage—first-gen versus continuing gen, URM versus non-URM, female versus male—and across med, vet, and APP student groups. Plus signs indicate means, horizontal lines within boxes indicate medians, boxes indicate interquartile ranges, and whisker lines indicate minimum and maximum ranges. One asterisk indicates P < .05, 2 asterisks indicate P < .01, and 3 asterisks indicate P < .001. Abbreviations: ACEs-14, Adverse Childhood Experience Scale Questionnaire; first-gen, first-generation college graduate; continuing gen, continuing generation college graduate; URM, underrepresented in medicine; med, medical; vet, veterinary; APP, advanced practice provider; ns, not significant.

ACEs and social disadvantage

To investigate the relationship between trauma exposure and social disadvantage, bivariate analyses between ACEs-14 scores and the demographic categories of female gender, first-gen status, and URM status were carried out (see Figure 1). For gender, there was no statistically significant difference in ACEs-14 scores for female versus male students (female: median = 1.0 [IQR = 0.0, 3.0], mean = 2.3 [SD = 2.73] vs. male: median = 1.0 [IQR = 0.0, 4.0], mean = 2.3 [SD = 2.76]; P = .85). For first-gen status, there was a highly statistically significant difference in ACEs-14 scores between first-gen and continuing gen students, with higher ACEs-14 scores for first-gen students (first-gen: median = 2.0 [IQR = 1.0, 5.0], mean = 3.6 [SD = 3.36] vs. continuing gen: median = 1.0 [IQR = 0.0, 3.0], mean = 1.6 [SD = 2.06]; P < .001). For URM status, there was also a highly statistically significant difference in ACEs-14 scores for URM versus non-URM students, with higher ACEs-14 scores for URM students (URM: median = 2.0 [IQR = 1.0, 5.0], mean = 3.3 [SD = 3.28] vs. non-URM: median = 1.0 [IQR = 0.0, 3.0], mean = 1.9 [SD = 2.37]; P = .002). Additional binary logistic regression analyses showed that URM and first-gen students had 56% (odds ratio [OR] = 1.56; P = .049) and 163% (OR = 2.63; P < .001), respectively, higher odds of having ≥ 4 ACEs-14 scores, compared to non-URM and continuing gen students.

ACEs and psychological distress

Next, we investigated the relationship between trauma exposure and psychological distress by analyzing MSWBI scores across a range of ACEs-14 scores. Figure 2 depicts the average psychological distress scores across 3 levels of ACEs exposure: 0, 1–3, and ≥ 4, where higher scores indicate greater psychological distress. These 3 groups were statistically significantly different, indicating that higher ACEs-14 scores were positively associated with higher psychological distress scores (P < .001). Using a linear regression analysis, we found a statistically significant unadjusted effect of ACEs on psychological distress. Specifically, an increase in the ACEs-14 score by 1 point was associated with an increase in the psychological distress score by 0.16 points (unadjusted beta = 0.16; P < .001; see Table 2). Further analyses showed that for every additional ACEs-14 reported, the odds of psychological distress, defined as MSWBI score ≥ 4, increased by 13% (OR = 1.13; P = .023).

Figure 2.

ACEs-14 scores and psychological distress (MSWBI) scores for medical, veterinary, and advanced practice provider students (n = 240), University of California Davis, July 2019. The distributions of MSWBI scores are depicted across levels of ACEs-14 scores using a box and whisker plot. Plus signs indicate means, horizontal lines within boxes indicate medians, boxes indicate interquartile ranges, and whisker lines indicate minimum and maximum ranges. Three asterisks indicate P < .001. Abbreviations: ACEs-14, Adverse Childhood Experience Scale Questionnaire; MSWBI, Medical Student Well-Being Index.

Table 2.

Unadjusted and Adjusted Associations of ACEs, Resilience, and Social Disadvantage With Psychological Distress for Medical, Veterinary, and Advanced Practice Provider Students (n = 240), University of California Davis, July 2019

Covariate	Unadjusted, beta (SE)	P value	Adjusted, beta (SE)	P value
ACEs-14 score	0.16 (0.04)	< .001	0.11 (0.05)	.036
Resilience score	−1.03 (0.18)	< .001	−1.07 (0.22)	< .001
URM status	0.15 (0.26)	.57	0.11 (0.29)	.72
Female gender	0.15 (0.28)	.77	0.11 (0.31)	.73
First-gen status	0.38 (0.30)	.21	0.26 (0.31)	.39

Abbreviations: ACEs, adverse childhood experiences; SE, standard error; ACEs-14, Adverse Childhood Experience Questionnaire; URM, underrepresented in medicine; first-gen, first generation college graduate.

ACEs and resilience

Next, we investigated the relationship between ACEs exposure and resilience. The overall distribution of resilience (i.e., BRS) scores showed that the majority of students reported resilience in the normal to high range (normal: 76%, 183/240; high: 10%, 25/240), with few students reporting low resilience (13%, 32/240). Expecting the negative association between ACEs and resilience reported in the literature, resilience scores were compared with ACEs-14 scores using linear regression analysis (see Figure 3).⁵⁴ There was a high number of students who had ≥ 4 ACEs-14 scores and normal or high resilience (88%, 53/60), while very few reported ≥ 4 ACEs-14 scores and low resilience (12%, 7/60). Overall, there was not a statistically significant relationship between ACEs-14 scores and resilience scores (P = .936).

Figure 3.

ACEs-14 scores and resilience (BRS) score levels for medical, veterinary, and advanced practice provider students (n = 240), University of California Davis, July 2019. The scatter plot showing resilience score by ACEs-14 score, with each point in the figure representing 1 or more students with that score. Dotted lines indicate the boundaries of published BRS norms⁵⁰ for low (1.00–2.99), normal (3.00–4.30), and high (4.31–5.00) resilience. Fit lines (gray lines) and 95% confidence intervals (gray shading) are plotted for reference. Abbreviations: ACEs-14, Adverse Childhood Experience Scale Questionnaire; BRS, Brief Resilience Scale.

ACEs, resilience, and psychological distress

To further investigate the relationship between ACEs, resilience, and social disadvantage on psychological distress, a full regression model was fit that included ACEs-14 scores, resilience scores, female gender, URM status, and first-gen status as predictors of psychological distress scores. In the adjusted full model, resilience and ACEs-14 scores were statistically significantly associated with psychological distress scores (ACEs-14: P = .036; resilience: P < .001), where an increase of 1 point in ACEs-14 score was associated with an increase in psychological distress score of 0.11 points (beta = 0.11), and an increase of 1 point in resilience score was associated with a decrease in psychological distress score of 1.07 points (beta = −1.07; see Table 2).

Analysis of the social disadvantage variables (i.e., female gender, URM status, and first-gen status) within the regression model failed to show a statistically significant association of social disadvantage with psychological distress (see Table 2). In addition, social disadvantage did not moderate the effect of ACEs on resilience (see Supplemental Digital Appendix 2 at [LWW INSERT LINK]), although we did find a small effect of social disadvantage on the relationship between ACEs and psychological distress. Specifically, URM students had a lower increase in psychological distress with increasing ACEs, compared to non-URM students (beta = −0.18, P = .042).

Discussion

This study found that students from some socially disadvantaged backgrounds reported statistically significantly higher ACEs-14 scores than their nondisadvantaged peers across 3 health professions schools at our institution. Specifically, URM and first-gen students had 56% and 163%, respectively, increased odds of reporting ≥ 4 ACEs. Adding to the emerging literature,^7,9,55,56 we found that medical and APP students at our institution had higher ACEs-14 scores than veterinary students. Considering the association between URM and first-gen status with higher ACEs exposure, this is likely related to differences in sociodemographic composition of the student body across the 3 programs. Notably, however, the higher ACEs prevalence among socially disadvantaged students coexisted with normal to high levels of resilience, even as ACEs were positively associated with psychological distress.

ACEs and social disadvantage

While it is heartening that URM and first-gen students in our study reported high levels of resilience at matriculation, the increased ACEs exposure they reported, compared to non-URM and continuing gen peers, may increase the risk of negative mental health and academic outcomes over the course of their education. Although we did not find a statistically significant relationship between socially disadvantaged background and psychological distress, we caution interpreting this as a negligible risk for these students. Psychological distress is likely at its lowest point at matriculation but may worsen for certain individuals depending on multiple factors over the course of health professions training. There is mounting evidence that inequalities in academic outcomes of URM medical students (data on other student groups are scant) are related to systemic and institutional factors, such as assessment practices, the learning environment, and differential learning experiences.⁵⁷ The interaction between these factors and ACEs may tip the balance between risk and protective factors toward vulnerability. Health professions educators and academic institutions have an the ethical imperative to assign adequate resources and implement efforts to eliminate structural or systemic inequalities that contribute to academic or mental health problems among students from disadvantaged backgrounds. Those resources and efforts may, for example, address unmet socioeconomic needs (e.g., food insecurity, housing) and problems in the learning environment (e.g., discrimination based on ethnic identity) in addition to addressing mental health or academic progression issues at the individual level. Focusing on psychological distress or differential academic attainment as issues pertaining to individual students, while ignoring their structural and social determinants, misses the root causes.

Social disadvantage and psychological distress

Our mixed findings pertaining to the relationship between social disadvantage and psychological distress add to the growing literature on mental health and well-being for socially disadvantaged groups, including female, URM, and first-gen medical students.^58–61 The need for further study is highlighted by a recent systematic review of studies on URM physicians, residents, and students that found that comparison in burnout between URM and non-URM is inconclusive with some studies reporting higher burnout in URM individuals, some finding no difference, and some finding that URM individuals have greater odds of recovering from burnout.⁶² We plan to look at the longitudinal course of psychological distress of students from socially disadvantaged backgrounds during the follow-up portion of this study, considering the literature showing that racial/ethnic and sexual orientation/gender identity minority backgrounds report high rates of mistreatment,³² microaggressions,⁶³ stereotype threat,⁶⁴ discrimination,⁶⁵ social isolation,⁶⁶ financial stress,⁶⁷ and academic difficulties^68,69 throughout their education. In the sole study that compared depression symptoms between the first and fourth year of medical school, the increase in depression among all students was statistically significantly related to the school’s negative racial climate.⁷⁰ This suggests that psychological distress may not be linked to disadvantaged status at matriculation, but may become more apparent as a function of students’ experiences within a school over time.

ACEs and resilience

Contrary to the literature,⁵⁴ higher ACEs-14 scores were not associated with lower resilience scores in our findings. We speculate that this is likely because our study included highly resilient students. Namely, there were very few students with high ACEs and low resilience, with most students across all ACEs levels scoring in the normal or high resilience range (see Figure 3). This could suggest a selection of highly resilient matriculants within the studied population.

Implications for equity- and trauma-informed health professions education and interventions

Our findings related to ACEs prevalence and the association of ACEs with psychological distress support current efforts toward trauma-informed education practices within health professions schools.¹⁴ In their scoping review, Burns and colleagues summarized 55 studies from several countries and identified the most common learning objectives and areas for improvement, such as the need for content on secondary trauma or occupational exposure to trauma.¹⁴ Skills commonly emphasized are screening for trauma and responding to trauma disclosures, while pedagogies include lectures and active learning exercises to practice skills.¹⁴ The authors highlighted the need to educate clinicians in “trauma stewardship,” which is the mindful practice of self-care, while caring for the suffering of others.¹⁴ This practice may be particularly challenging for students with a history of ACEs, as suggested by studies that found an association between ACEs and increased risk of depression and burnout in physicians⁷¹ and ACEs and occupational impairment in health care workers.⁷²

The implementation of trauma-informed approaches to health professions education is congruent with other contemporary initiatives that challenge the emphasis on individual failure and deficits as drivers of inequities in educational outcomes. Collectively, they emphasize instead the role of the learning environment and other institutional and systemic factors, such as social and structural determinants of health,^73,74 social justice,^75,76 and health equity.^77,78 Elsewhere, we applied a social determinants of health lens to a framework for the study of resilience in health professions education.⁷⁹ Others have proposed similar frameworks articulated around social determinants, risk factors, and needs⁸⁰ and have put such frameworks into practice in at least one institution.⁸¹

Limitations

Our results represent a cross-sectional analysis of matriculation data in a longitudinal study of health professions students. The associations we found at matriculation between ACEs and psychological distress and between resilience and psychological distress do not imply causality and cannot be assumed to apply to later years of study. Since a decrease in resilience scores across the curriculum has been documented in some studies,⁸² it is possible that the detrimental effect of ACEs on resilience may become more evident as students are challenged over the course of their programs of study.

Our findings correspond to 3 health professions schools at the University of California Davis and cannot be assumed to apply to other groups of health professions students (e.g., pharmacy students) or other academic institutions. Also, our study did not include information on other socially disadvantaged backgrounds, such as sexual orientation, gender expression, and family SES. Studies powered to allow intersectional analysis of multiple socially disadvantaged identities may show a greater impact on students’ psychological distress than we saw in our population. Finally, this analysis does not include positive childhood experiences. Multiple positive childhood experiences have been reported to be common even in populations with high exposure to ACEs.^83,84 The fact that students in our study who report high ACEs also report resilience scores in the normal to high range hint at the presence of undetected positive childhood experiences. Such experiences may be associated with successful educational attainment and future careers.

Conclusions

Despite the limitations, our findings are consistent with an important body of data on ACEs and health and educational inequities. Cross-sectional, longitudinal, and intervention studies (e.g., participation in an early childhood education program) have shown a graded relationship between ACEs exposure and educational attainment (including higher education) in youth and adults.^85–87 Future prospective research across institutions may further elucidate the interaction between structural and social determinants of inequities in mental health and academic progression, thereby providing the evidence needed for the development of interventions aimed at the elimination of such inequities.

Supplementary Material

Supplemental digital content for this article is available at [LWW INSERT LINK].