Adverse childhood experiences are associated with at-risk drinking, cannabis, and illicit drug use in females but not males: An Emergency Department study

Abstract

Background:

Adverse childhood experiences (ACEs) are associated with adult substance use in the general population. Given pervasive health disparities among underserved populations, understanding how ACEs are associated with substance use among urban Emergency Department (ED) patients could help inform design of effective screening, brief interventions, and referral to treatment.

Objectives:

To estimate gender differences in prevalence of separate and cumulative ACEs among a sample of urban ED patients, and assess its association with at-risk drinking (4+/5+ drinks for females/males), cannabis, and illicit drug use. We hypothesized that the association between ACEs and each outcome would be stronger among females than males.

Methods:

Cross-sectional survey data were obtained from 1,037 married/partnered ED patients (53% female) at a public safety-net hospital. Gender-stratified logistic regression models were estimated for each substance use outcome.

Results:

One+ ACEs were reported by 53% of males and 60% of females. Females whose mother was a victim of domestic violence had greater odds of at-risk drinking compared to females who did not report this ACE (AOR=1.72; 95% CI 1.03, 2.88). Females’ cumulative ACEs were associated with cannabis use (OR=2.26, 95% CI 1.06, 4.83) and illicit drug use (OR=3.35; 95% CI 1.21, 9.30). Males’ separate and cumulative ACEs were not associated with increased likelihood for any of the outcomes.

Conclusion:

ACEs are associated with greater odds of substance use among female than male ED patients. The prevalence of ACE exposure in this urban ED sample underscores the importance of ED staff providing trauma-informed care.

1. Introduction

Substance use is a major contributor to morbidity and mortality. Heavy alcohol use, for example, is associated with health and social harms to self (Rehm et al., 2006) and others (Karriker-Jaffe et al., 2017). Trends towards legalization of recreational cannabis are expected to result in increased car crashes, fatalities and injuries, and Emergency Department (ED) presentations (Hall & Lynskey, 2016). Illicit drug use is related to acute toxicities (Liakoni et al., 2015) and an array of medical and mental health conditions (Han et al., 2010).

Persons exposed to adverse childhood experiences (ACEs) are more likely to have numerous health and behavioral problems in adulthood, including substance use (Campbell et al., 2016; Hughes et al., 2017). An analysis of the National Epidemiologic Survey of Alcohol and Related Conditions (NESARC) showed that physical, emotional, and sexual abuse and physical and emotional neglect were associated with abuse or dependence of alcohol, cannabis, and illicit drugs, and misuse of prescription drugs (Afifi et al., 2012). Anda and colleagues (2006) summarized the neurobiological and epidemiological evidence by which early life stress (such as abuse and other adverse experiences) can cause lasting brain dysfunction in hippocampus, amygdala, media prefrontal cortex, and other limbic structures that may mediate anxiety and mood dysregulation following early childhood abuse. The resulting neurobiological disruptions can be linked to stress reactivity, lowered impulse control, and impaired cognitive function (Lovallo et al., 2013). In turn, these mechanisms may contribute to the likelihood that those exposed to ACEs will engage in harmful substance use behaviors. Neurobiological disruptions are also independently linked to substance use behaviors, and ACEs may exacerbate harmful substance use associations.

Research exploring gender differences in the association between ACE and adult substance use has produced mixed results. For example, an analysis of a large representative Canadian sample did not find strong evidence that gender moderated the association between ACE and alcohol and drug dependence (Fuller-Thomson et al., 2016). In contrast, gender differences were observed in an analysis of over 7,000 U.S. Kaiser Permanente members; women but not men who reported childhood emotional abuse were at elevated risk for self-reported alcohol problems (Strine et al., 2012). When cumulative ACEs score of 4+ was considered, however, the strength of the association between 4+ ACEs and alcohol problems was greater for men than for women (adjusted Odds Ratios [AORs] 9.9 vs. 2.7) (Strine et al., 2012). In an analysis of gender differences in the effects of ACE on alcohol, drug, and polysubstance-related disorders based on Wave 2 of the NESARC data, Evans and colleagues (2017) reported that increased risk for each type of substance use disorder was associated with more exposure to ACEs, and women had lower risk than men. Moderation analysis, however, showed that with increasing ACE exposure, the gender gap for alcohol use disorder narrowed, especially for women with exposure to 3+ ACEs. For drug use disorders, moderation effects showed that women with 3+ ACEs were at greater risk than men with the same exposure. For polysubstance-related disorder, the gender gap increased such that men exposed to 3+ ACEs were at greater risk than women with the same exposure (Evans et al., 2017). Among adolescents, a recent study found gender differences in the effects of ACEs on substance use and delinquency and showed that ACEs were related to substance use for girls but not boys (Leban & Gibson, 2020). Some researchers have theorized that ACE exposure results in women becoming more sensitized than men to the effects of adult stressors due to impaired stress-response coping mechanisms and decreased emotional regulation, thereafter heightening women’s vulnerability to substance use disorders (McLaughlin et al., 2010; Myers et al., 2014; Young-Wolff et al., 2012).

To date, few studies have examined the prevalence of ACEs among ED populations and its association with substance use outcomes. Rothman and colleagues (2011) found that females had higher ACE scores than males among a sample of urban ED patients ages 14–21 years, but ACEs were related to drinking style (e.g., hazardous drinking, drinking to cope) among both. Among a sample of pregnant urban ED patients, Nelson et al. (2010) found that 29% reported at least one episode of childhood physical violence, and 14% reported childhood rape; both groups were more likely to be hazardous drinkers than those without these ACE exposures. Additional research on ACE and substance use among ED samples is warranted for several reasons. First, alcohol, cannabis, and illicit drug use are more prevalent among ED patients than in the general population (Cherpitel & Yu, 2012; D’Onofrio et al., 2006). Given that ACEs are related to increased likelihood of substance use in the general population, it is important to quantify these associations among ED populations in which the impact of poverty and other barriers to health care may exacerbate disparities in health outcomes (Wade Jr. et al., 2016). Second, most patients seeking non-emergency care at urban EDs are Medicaid recipients, uninsured or underinsured and not likely to have a primary care provider or may not access health care otherwise (Murnik et al., 2006; Tang et al., 2010). Consistent with the importance of providing trauma-informed care (Fischer et al., 2019; Levy-Carrick et al., 2019), having more information on ACE and substance use outcomes could help inform the design of effective screening, brief interventions, and referral to treatment (SBIRT) approaches to alcohol, tobacco and other drug use among ED patients (Cunningham et al., 2009).

The goal of this study is to analyze gender differences in prevalence of separate and cumulative ACEs, and to determine their association with at-risk drinking, cannabis use and illicit drug use among a sample of married/partnered urban ED patients. We hypothesized that the associations between ACEs and each substance use outcome would be stronger among females than males. As part of our multivariate model building strategy, we include two psychosocial covariates that have been shown to be related to ACEs and substance use, impulsivity (Lovallo et al., 2013; Winhusen & Lewis, 2013) and stressful life events (Armstrong et al., 2018). The models also include covariates representing spouse/partner substance use since previous analysis of the data demonstrated significant associations between the smoking behaviors of study participants and their spouse/partners (Cunradi et al., 2019).

2. Methods

2.1. Study design and setting

Survey data were collected as part of a cross-sectional study on drinking, drug use and intimate partner violence (IPV) among a sample of married/partnered ED patients at an urban Level I trauma center in Northern California (Caetano et al., 2019). Previous analysis showed a substantial proportion of the study’s sample reported indicators of low socioeconomic status. Nearly one third had not completed high school; less than 10% had graduated from college. Almost half reported that they sometimes or often ran out of food during the past 12 months and didn’t have enough money to get more (Cunradi et al., 2020). The hospital where the study was conducted is part of a county-wide integrated public health care system. Its ED has an annual census of 72,000 and serves as the county’s safety-net provider. Most (61%) patients are covered by Medicaid; 17% of patients are uninsured. Most patients are African American or Hispanic. The hospital’s Institutional Review Board approved the project.

2.2. Subject selection

Study eligibility criteria were: 18–50 years old; English or Spanish speaker; resident of the county in which the hospital is located; and married, cohabiting, or in a romantic (dating) relationship for the past 12 months. Patients who were intoxicated, experiencing acute psychosis or suicidal or homicidal ideation, were cognitively/psychologically impaired and unable to provide informed consent, in custody by law enforcement, or in need of immediate medical attention (i.e., Emergency Severity Index [ESI] level 1 or 2) (Gilboy et al., 2011) were ineligible and excluded. Data collection was conducted from February through December 2017. Due to staffing constraints, we did not seek to proportionately recruit participants from all ED shifts. Instead, 2 interviewers per shift staffed the ED during weekday peak volume hours (9am – 9pm) to recruit eligible participants to the study. Patients could opt to be interviewed in English or Spanish. For the latter, a Spanish version of the questionnaire, which had been validated through translation into Spanish and retranslation into English followed by verification, was then used. Those who chose to be interviewed in Spanish were interviewed by bilingual Research Assistants (RAs).

Figure 1 shows the recruitment sequence. The RAs identified potentially eligible participants through the ED’s electronic patient information system. Most patients were there to receive non-emergency medical care (e.g., ESI 4 or 5) or to renew prescription medication. In all cases, patients were stabilized before being interviewed. The RAs located and conducted face-to-face screening with patients in the ED waiting room or in a treatment cubicle. Eligible participants were offered the opportunity to participate in a confidential, face-to-face survey interview for which they would receive a $30 grocery store gift card incentive. The RAs obtained informed consent in a private area adjacent to the ED waiting room, or in the subject’s room without others present. Twenty-nine participants terminated the survey interview before completion, due primarily to interruption for medical services (e.g., patient transported to ultrasound or X-ray). Thus, 1,037 participants (53% female) completed the survey interview. Survey data were collected by the RAs using computer assisted personal interview (CAPI) techniques with tablets running the Qualtrics platform. Average survey interview completion time was 37 minutes (SD 20.7). The multivariate analysis herein is based on complete data from 999 participants.

Figure 1:

Study sample recruitment

2.3. Measurements

Exposure to adverse childhood experiences.

Due to survey time constraints, we measured adverse childhood experiences with a brief version of the ACE scale. This version measures exposure to six adverse experiences the patient may have had “while they were growing up during their first 18 years of life:” (1) exposure to a mentally ill person in the home; (2) parent/caregiver alcoholism; (3) sexual abuse; (4) physical abuse; (5) psychological abuse; and (6) violence directed against the respondent’s mother (Cabrera et al., 2007). Each exposure was measured with one question. In contrast, the ACE as developed by Felitti et al. (1998) measured exposure with 2–4 questions per category, and contained an additional category (household criminal behavior) measured with one question. Based on the brief scale, we created six dichotomous variables that represented exposure to each separate ACE. We next created a 4-level variable to represent exposure to 0, 1, 2, and 3 or more ACEs. Cronbach’s α for the brief ACE scale was 0.74.

2.3.1. Outcome measurements

We created variables for three substance use outcomes. For at-risk drinking, we asked participants about their past year drinking patterns, including greatest number of drinks in one day. A “drink” was defined as a 12-ounce can of beer, a 5-ounce glass of wine, or a 1-ounce shot of liquor. Females who reported that they drank at least 4 drinks, and males at least 5 drinks, were classified as at-risk drinkers (National Institute on Alcohol Abuse and Alcoholism, 2018). For cannabis use, we asked participants, “How many times during the past 12 months, or 365 days, did you use marijuana or hashish (weed, pot, hash) without a doctor’s instruction?” A dichotomous variable was created representing any past-year cannabis use. Recreational cannabis use was legalized in California for those 21+ years old in November 2016. We next asked participants if they used any illicit drugs in the past year (i.e., cocaine, heroin, amphetamines, opioid prescription drug misuse). A dichotomous variable was created representing any past-year illicit drug use.

2.3.2. Covariates

Sociodemographic factors.

Participants self-reported their gender. Three patients identified as transgender; due to small numbers, these cases were excluded from the analysis. We created a dichotomous male/female variable. Participants were asked to name the racial or ethnic group(s) that best describes them. Those who selected more than one category were categorized as multiethnic. For the analyses, these groups were recoded into a 4-category race/ethnicity variable: Hispanic; Black; other; and white. Participant age was used as a continuous variable.

Psychosocial variables.

Impulsivity was measured with a 3-item scale that assessed respondents’ agreement with the following statements: I often act on the spur of the moment without stopping to think; You might say I act impulsively; Many of my actions seem to be hasty (Caetano et al., 2000). Items were scored from 1–4, with a higher score representing greater impulsivity. Cronbach’s α = .79. Past-year stressful life events (e.g., laid off from a job; death of family member or close friend) were measured with a 14-item scale taken from the AUDADIS-IV (Ruan et al., 2008). The items created an index that varied from 0 to 14. Cronbach’s α = .73.

Spouse/partner substance use.

We used the 3-item AUDIT-C (Alcohol Use Disorders Identification Test-Consumption) to measure the participant’s assessment of his/her spouse/partner’s drinking (Bradley et al., 2003; Bush et al., 1998). Sum of scores for the 3 items range from 0 to 12, and male/female partners with a score above 4 or 3, respectively, were considered hazardous drinkers. Cronbach’s α = .81. We asked participants if their spouse/partner used marijuana or any illicit drugs (i.e., cocaine, heroin, amphetamines, opioid prescription drug misuse) in the past year. Dichotomous variables were created representing past-year spouse/partner cannabis use and illicit drug use, respectively.

2.4. Statistical analyses

Analyses were conducted in IBM SPSS v. 25. We used chi square tests of independence to analyze gender differences in categorical participant characteristics, exposure to cumulative ACEs (1, 2, 3+ ACEs) and each ACE scale item, and t-tests for continuous variables. We used the Bonferroni correction to account for the multiple tests performed (p<0.003). Next, using a gender-stratified approach, we calculated unadjusted and adjusted Odds Ratios [ORs] and 95% Confidence Intervals [CIs] for the association between each ACE item and at-risk drinking, cannabis use, and illicit drug use. Adjusted odd ratios accounted for age, race/ethnicity, impulsivity, stressful life events, and partner’s substance use. We conducted chi square tests to analyze the bivariate associations between cumulative ACEs and each substance use outcome. Lastly, we estimated gender-stratified logistic regression models for cumulative ACEs and each substance use outcome. Each model was adjusted for age, race/ethnicity, impulsivity, stressful life events, and spouse/partner’s substance use. Missing data ranged from 0 – 1.6% for the variables in the study and were dropped from the analysis through listwise deletion. Given the low proportion of missingness, it is unlikely that data imputation would improve estimations (Allison, 2001). Bivariate analysis of the dependent and independent variables found that only impulsivity significantly differed between complete and incomplete cases.

3. Results

Sample characteristics are shown in Table 1. A greater proportion of males than females had engaged in at-risk drinking (34.3% vs. 20.4%) and in cannabis use (30.5% vs. 23.9%). Approximately 53% of males and 60% of females reported one or more ACEs; cumulative ACE exposure did not differ significantly by gender. A greater proportion of females than males reported that they grew up with a depressed or mentally ill household member (29.2% vs. 18.2%) or experienced sexual abuse (20.4% vs. 6.6%). A greater proportion of white males and males in the “other” racial/ethnic group reported 3+ ACEs, and a smaller proportion of Hispanic males reported 3+ ACEs. Similar patterns were observed among females (data not shown).

Table 1.

Sample Characteristics

	Males	Females
	(n=484)	(n=550)
Age (Mean, SD)*	36.5 (8.2)	34.0 (8.5)
Race/ethnicity
Black	26.2	31.3
Hispanic	52.9	47.6
Other	14.5	14.5
White	6.4	6.5
Stressful life events (Mean, SD)	3.7 (2.8)	3.3 (2.7)
Impulsivity (Mean, SD)	5.3 (2.5)	5.4 (2.6)
At-risk drinking*	34.3	20.4
Cannabis use	30.5	23.9
Illicit drug use*	17.0	8.1
Spouse/partner problem drinking	19.4	22.5
Spouse/partner cannabis use*	18.0	26.2
Spouse/partner illicit drug use	4.6	6.9
Adverse childhood experiences
0	47.1	40.3
1	22.5	20.3
2	14.7	15.5
3+	15.7	23.9
Depressed/mentally ill household member*	18.2	29.2
Problem drinker in household	38.7	40.6
Psychological/emotional abuse	16.6	21.5
Physical abuse	15.2	14.9
Mother victim of domestic violence	18.8	22.7
Sexual abuse*	6.6	20.4

^*Significant gender difference (p < 0.003)

Table 2 shows the unadjusted and adjusted ORs for the association of each ACE scale item with the 3 substance use outcomes. Among males, none of the associations remained significant after adjustment for covariates. Among females, those who stated that their mother was a victim of domestic violence were more likely to report at-risk drinking compared to females who did not report this ACE (AOR=1.72; 95% CI 1.03, 2.88; p<0.05).

Table 2.

Associations between exposure to each adverse childhood experience and substance use outcomes

	At-risk drinking		Cannabis use		Illicit drug use
	ORunadjusted	AOR	ORunadjusted	AOR	ORunadjusted	AOR
	(95%CI)	(95%CI)	(95%CI)	(95%CI)	(95%CI)	(95%CI)
Depressed/mentally ill household member
Men	0.79 (0.46, 1.36)	0.85 (0.49, 1.47)	2.62 (1.50, 4.56)b	1.22 (0.63, 2.35)	1.55 (0.84, 2.89)	0.90 (0.44, 1.81)
Women	1.72 (1.06, 2.79)a	1.42 (0.85, 2.36)	2.53 (1.55, 4.15)c	0.76 (0.42, 1.40)	3.14 (1.59. 6.20)b	1.75 (0.80, 3.83)
Problem drinker in household
Men	0.90 (0.58, 1.39)	0.95 (0.62, 1.46)	1.35 (0.86, 2.13)	1.28 (0.74, 2.22)	2.02 (1.18, 3.44)a	1.34 (0.75, 2.41)
Women	1.40 (0.87, 2.23)	1.15 (0.72, 1.84)	1.49 (0.93, 2.39)	0.83 (0.48, 1.45)	2.31 (1.17, 4.56)a	2.07 (0.97, 4.42)
Psychological/emotional abuse
Men	0.90 (0.45, 1.42)	0.70 (0.39, 1.25)	1.69 (0.95, 3.00)	0.71 (0.34, 1.45)	1.34 (0.69, 2.58)	0.72 (0.35, 1.50)
Women	1.25 (0.73, 2.13)	1.18 (0.69, 2.03)	1.63 (0.95, 2.77)	0.82 (0.43, 1.56)	2.41 (1.22, 4.79)a	1.75 (0.78, 3.95)
Physical abuse
Men	1.25 (0.70, 2.24)	1.10 (0.62, 1.96)	1.69 (0.94, 3.04)	0.97 (0.47, 2.00)	1.88 (0.99, 3.59)	1.60 (0.80, 3.21)
Women	1.24 (0.65, 2.35)	0.97 (0.51, 1.82)	1.68 (0.89, 3.18)	0.55 (0.25, 1.20)	1.75 (0.77, 3.97)	0.93 (0.34, 2.54)
Mother victim of domestic violence
Men	0.77 (0.44, 1.36)	0.59 (0.34, 1.02)	1.30 (0.73, 2.30)	1.28 (0.64, 2.54)	2.29 (1.24, 4.22)a	1.31 (0.68, 2.52)
Women	1.90 (1.13, 3.20)a	1.72 (1.03, 2.88)a	2.15 (1.27, 3.64)b	1.06 (0.57, 1.99)	2.79 (1.42, 5.49)a	1.62 (0.74, 3.57)
Sexual abuse
Men	1.46 (0.65, 3.27)	1.39 (0.64, 3.01)	1.12 (0.49, 2.57)	1.00 (0.37, 2.71)	2.06 (0.88, 4.84)	2.11 (0.85, 5.24)
Women	0.88 (0.49, 1.58)	0.71 (0.39, 1.28)	1.15 (0.64, 2.06)	1.06 (0.56, 2.00)	1.32 (0.61, 2.87)	1.20 (0.51, 2.83)

AORs adjusted for age, race/ethnicity, impulsivity, stressful life events, and partner’s substance use.

^aP < 0.05

^bP < 0.01

^cP < 0.001

Bivariate analysis (Table 3) shows that among males, the proportion of those categorized as at-risk drinkers did not vary by number of cumulative ACEs, whereas the proportion who used cannabis or illicit drugs varied by number of cumulative ACEs. Among females, there were significant differences between the proportion who engaged in each substance use outcome and number of cumulative ACEs.

Table 3.

Bivariate associations between cumulative ACEs and substance use outcomes

	Males			Females
At-risk Drinking
No. of ACEs	No	Yes	Chi-square	No	Yes	Chi-square
0	67.1	32.9	0.54	87.3	12.7	14.96
1	63.3	36.7		77.5	22.5	p=.002
2	66.2	33.8		74.1	25.9
3+	64.5	35.5		71.8	28.2
Cannabis Use
No. of ACEs	No	Yes	Chi-square	No	Yes	Chi-square
0	75.6	24.4	14.98	86.8	13.2	23.56
1	70.6	29.4	p=.002	71.6	28.4	p<.001
2	67.1	32.9		69.0	31.0
3+	52.0	48.0		66.7	33.3
Illicit Drug Use
No. of ACEs	No	Yes	Chi-square	No	Yes	Chi-square
0	89.3	10.7	21.65	96.8	3.2	29.87
1	84.4	15.6	p<.001	94.5	5.5	p<.001
2	77.5	22.5		92.9	7.1
3+	67.1	32.9		80.8	19.2

Results of the multivariate analysis between cumulative ACEs and each substance use outcome (Table 4) showed distinct gender differences. Among men, ACEs were not associated with any of the outcomes. Instead, impulsivity (AOR=1.10; 95% CI 1.01, 1.20) and having a spouse/partner who was a hazardous drinker (AOR= 2.61; 95% CI 2.61, 4.37) were positively associated with at-risk drinking. Stressful life events (AOR=1.16; 95% CI 1.05, 1.28) and having a spouse/partner who used cannabis (AOR=8.99; 95% CI 4.62, 17.50) were positively associated with male’s cannabis use. Hispanic males had lower odds of cannabis use compared to white males (data not shown). Impulsivity (AOR=1.19; 95% CI 1.07, 1.33), stressful life events (AOR=1.19; 95% CI 1.07, 1.32) and having a spouse/partner who used illicit drugs (AOR=19.01; 95% 5.47, 66.01) were positively associated with male’s illicit drug use.

Table 4.

Multivariate associations between ACEs and substance use outcomes.

	Males	Females
	OR (95% CI)	OR (95% CI)
At-risk drinking:
No. of ACE: (ref: none)
1	1.01 (0.61, 1.69)	1.34 (0.70, 2.57)
2	0.85 (0.46, 1.57)	1.70 (0.87, 3.35)
3+	0.73 (0.38, 1.40)	1.41 (0.75, 2.68)
Impulsivity	1.10 (1.01, 1.20)*	1.06 (0.97, 1.16)
Stressful life events	1.03 (0.95, 1.12)	1.12 (1.02, 1.22)*
Partner hazardous drinking	2.61 (1.56, 4.37)***	4.23 (2.60, 6.87)***
Cannabis use:
No. of ACE: (ref: none)
1	0.84 (0.42, 1.69)	1.45 (0.69, 3.04)
2	0.80 (0.36, 1.79)	2.26 (1.06, 4.83)*
3+	1.40 (0.65, 3.03)	0.76 (0.35, 1.64)
Impulsivity	1.01 (0.91, 1.12)	1.08 (0.98, 1.20)
Stressful life events	1.16 (1.05, 1.28)**	1.30 (1.17, 1.45)***
Partner cannabis use	8.99 (4.62, 17.50)***	9.22 (5.19, 16.38)***
Illicit drug use:
No. of ACE: (ref: none)
1	1.03 (0.48, 2.20)	1.60 (0.47, 5.40)
2	1.20 (0.51, 2.80)	1.98 (0.58, 6.84)
3+	1.71 (0.77, 3.82)	3.35 (1.21, 9.30)*
Impulsivity	1.19 (1.07, 1.33)**	1.25 (1.09, 1.43)**
Stressful life events	1.19 (1.07, 1.32)**	0.99 (0.86, 1.14)
Partner illicit drug use	19.01 (5.47, 66.01)***	11.48 (4.68, 28.17)***

All models are adjusted for race/ethnicity and age.

^*p < .05;

^**p < .01;

^***p < .001

Among females, cumulative ACEs were not related to at-risk drinking; instead, stressful life events (AOR=1.12; 95% CI 1.02, 1.22) and spouse/partner hazardous drinking (AOR=4.23; 95% CI 2.60, 6.87) were positively associated with this outcome. Females who reported 2 ACEs were more likely to have used cannabis compared to those who did not report any ACE (AOR=2.26; 95% CI 1.06, 4.83). Stressful life events (AOR=1.30; 95% 1.17, 1.45) and having a spouse/partner who used cannabis (AOR=9.22; 95% CI 5.19, 16.38) were also positively associated with cannabis use. Lastly, for illicit drug use, females who reported 3+ ACEs were three times more likely to have used illicit drugs compared to females who reported no ACE (AOR=3.35; 95% CI 1.21, 9.30). Impulsivity (AOR=1.25; 95% CI 1.09, 1.43) and having a spouse/partner who used illicit drugs (AOR=11.48; 95% CI 4.68, 28.17) were also positively associated with illicit drug use. Hispanic females had lower odds for all substance use outcomes compared to white females. Black females had lower odds of illicit drug use, and women categorized in the “other” racial/ethnic group had lower odds of at-risk drinking, compared to white females (data not shown).

4. Discussion

Among an urban ED sample, most male (53%) and female (60%) participants reported that they had experienced at least 1 ACE. This is similar to the prevalence of ≥ 1 ACE observed among a racially and socially diverse urban sample in Philadelphia (Wade Jr. et al., 2016). In comparison, approximately 44% of Wave 2 NESARC respondents reported at least 1 ACE (Myers et al., 2014). In terms of exposure to each ACE scale item, a greater proportion of females than males reported growing up with a depressed or mentally ill household member and experiencing sexual abuse. These gender differences are in accord with those observed in large Canadian and American surveys (Meng & D’Arcy, 2016; Strine et al., 2012). Although there were significant bivariate associations between males’ cumulative ACEs and cannabis use and illicit drug use, these associations were not significant in the multivariate analyses. The overall picture that emerges is that ACEs appear to be more strongly related to substance use among females than males, thus confirming our gender-specific hypothesis. For example, among females, those whose mother experienced domestic violence were more likely to be at-risk drinkers compared to those not exposed to this ACE. Similarly, cumulative ACEs exposure was associated with females’ cannabis and illicit drug use. The gender differences seen in the results may reflect the greater impact of ACEs on females’ disinhibition (an aspect of impulsivity) leading to increased substance use (Nederkoorn et al., 2009; Winhusen & Lewis, 2013); ACEs may also result in greater stress sensitization among females which increases vulnerability to adult stressors, thereafter leading to substance use (Young-Wolff et al., 2012). Considering the dearth of ED-based studies on ACEs, gender and substance use, additional research is needed to further elucidate these associations among underserved populations.

4.1. Clinical implications

Given the prevalence of ACE exposure among males and females in this urban ED sample, the findings underscore the importance of ED staff providing trauma-informed care. This approach takes into consideration the lived experience of past and present psychological and physical trauma in patients’ lives, and how the patient may perceive and react to medical care (Fischer et al., 2019; Purkey et al., 2018). Briefly, the trauma-informed care approach encompasses trauma awareness and acknowledgment; safety and trustworthiness; choice, control, and collaboration; strengths-based and skills-building care; and cultural, historical, and gender issues (Purkey et al., 2018).

Behavioral interventions that promote mindfulness and mindfulness-based stress reduction may be effective in helping ED patients with ACE exposure reduce substance use behaviors that are linked with impulsivity and reactions to stressful life events (Brett et al., 2018; Oshri et al., 2018). Lastly, given the linkages between study participants’ substance use and that of their spouse/partner, treatment options that include the patient’s spouse/partner (including dual problem couples) may provide an added benefit compared to individual-focused treatment (McCrady et al., 2019; O’Farrell et al., 2017; Schumm et al., 2012; Schumm et al., 2019).

4.2. Strengths and limitations

To our knowledge, this study is among the first to analyze gender differences in ACE among an urban ED sample, and to analyze the relationship between ACE and at-risk drinking, cannabis, and illicit drug use after accounting for potentially confounding factors. The sample’s approximately equal numbers of males and females allowed for gender-stratified models of the substance use outcomes. Because significant health disparities characterize urban ED populations, this study addresses a gap in the current literature concerning the interrelationships between ACE and substance use in an underserved ED sample.

Study findings should be considered in the context of several limitations. First, the cross-sectional design limits causal inferences regarding the study’s observed associations. Second, the sample was obtained from a single urban ED, which may limit generalizability. Additionally, because the eligibility criteria specified that participants be ages 18–50, English or Spanish speakers, and be married, cohabiting, or in a dating relationship, the results herein may not be generalizable to those older than age 50, speakers of other languages, or who are not currently in a romantic partnership. Third, recall bias may have affected patients’ estimation of events over the previous 12 months. Fourth, use of the brief ACE scale may have underestimated exposure. Moreover, no information was collected about age at exposure. There is evidence, for example, that the type of ACE exposure during certain developmental stages may have varying neurobiological impacts that could negatively affect adult psychopathology (Schalinski et al., 2016).

4.3. Conclusions

Among a sample of urban ED patients, more than half of males and females reported one or more ACEs. Cumulative ACEs were associated with greater odds of cannabis and illicit drug use among females than males. Additionally, analysis of separate ACE exposures found that females whose mother was a victim of domestic violence had greater odds of at-risk drinking compared to females without this ACE; no significant associations, however, were observed among males. Because ACEs are linked with an array of social, behavioral and physical health problems throughout the lifespan and are associated with the intergenerational transmission of violence and adversity (Anda et al., 2006), reducing and preventing ACEs is a public health priority. Given the evidence for ACE-related health disparities among urban, low-income populations (Burke et al., 2011; Jimenez et al., 2017; Wade Jr. et al., 2016) the development of effective primary prevention should be the focus of multisector collaborative efforts among community, state and federal entities. Meanwhile, EDs may be effective partners in secondary prevention with enhanced identification and linkage to care for socially disadvantaged patients.